BMI

05-10-2023
----------------------------

05-10-2023
-------------------------------------
1.Audio not working properly in live Ebrainvoice version.2 application
2.Changed Ebrainvoice version.2 machine pipeline module for accommodating ML and DL modules and decision window size added in machine pipeline module.
3.Save resampling path in offline Ebrainvoice version.2 application

10-10-2023
-----------------------------------------------------------------------------------------------------------------------------------------
1.Design and development of EEG Wavelet Analyzer module started 45% completed
2.Development of image split module for DL
3.DL module integrating time an error occurred in execute pipeline and fixed(100% Completed)

11-10-2023
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.The error in the image split module has been resolved. An empty validation folder was generated in the image split module, and both training and test data were saved in the same folders. Both issues have been resolved and committed to SVN.
2.The design of the EEG Wavelet Analyzer UI module has been completed, and development has started. The coding for the EEG Wavelet Analyzer UI module is finished, but the implementation of the backend code is pending.
3.Parallely training and testing V5 voice absent data with Catboost classifier

12-10-2023
-------------------------------------------------------------------------
1.Channel-wise plotting of DWT and CWT coefficients and scalograms has been completed

13-10-2023
---------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Error in the Wavelet analyzer has been resolved
2.Training and testing voice absent EEG epoch data with catboost classifier

1. In EEG Data Viewer, if offline data is plotted, and then an attempt is made to plot live EEG data, the previous plot still remains. The offline EEG data plot is not cleared when we select real-time data. This error has been resolved
2. Integration of the training protocol code into the machine learning pipeline has been completed

02-11-2023
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1. In EEG Data Viewer, if offline data is plotted, and then an attempt is made to plot live EEG data, the previous plot still remains. The offline EEG data plot is not cleared when we select real-time data. This error has been resolved
2. Integration of the training protocol code into the machine learning pipeline has been completed

08-11-2023
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.I have updated the live protocol code to include both 'Repeat Training' and 'Repeat Once Training' options. If the user selects 'Repeat Training,' the beep will play continuously; otherwise, the beep will stop after the desired time. An issue was encountered in 'Repeat Training' mode, specifically a stopping issue, which is now being resolved in live prediction
2.I have resolved a bug related to feature extraction in the feature pipeline module

1. The preprocessing module was integrated with the live prediction module. However, after the first successful execution, the code started throwing the error "RuntimeError: main thread is not in the main loop" followed by "Tcl_AsyncDelete: async handler deleted by the wrong thread." I am currently working on resolving these issues.
2. Pending task: Issue with repeat training mode not solved. Problem: Module beeps continuously after it stops (pending task)

10-11-2023
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.The error "RuntimeError: main thread is not in the main loop" followed by "Tcl_AsyncDelete: async handler deleted by the wrong thread" was solved in live Mime Me (MindMe) software by adding the mtTkinter package to the filter UI, hardware UI, and resampling UI.
2.I explored and implemented a custom Tkinter UI in the MimeMe application. I changed the resampling UI framework from the Tkinter package to my custom Tkinter. After that, I integrated the resampling UI with the MimeMe software. Now, the issue is that the custom Tkinter UI is not popping up. I am resolving this issue. After solving this issue, I will test the resampling UI with live prediction to see if a thread issue occurs.
3. Pending task: Issue with repeat training mode not solved. Problem: Module beeps continuously after it stops (pending task)
4.Write live stream data into a CSV file and cross-check whether new data is being received from LSL correct or not
5.Explore and implement LSL buffers in Python. How to store live stream data, clear it, update it with new stream data, and explore the maximum amount of data the buffer can hold.
6.Set the clock according to the decision window used in the UI, and change the live stream processing algorithm according to these changes.

13-11-2023
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1."The live stream data-saving functionality has been implemented and tested using the live module."
2."Implemented the plotting of the saved stream in an offline EEG data viewer to validate the presence of eye blinks in the stream."
3."Resolved Custom Tkinter issues in the live prediction module."

14-11-2023
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
leave Reason:I want to visit IOB Bank to claim the approved subsidy amount from the central government and sign some documents.
15-11-2023
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Explore and study the OpenBCI GUI settings from the open bci document, and collaborate with team members to resolve the current visualization issue of OpenBCI
2."Implement a delay at the end of the beep in the live prediction module if 'enable training' is selected."
3"Solved the issue related to stopping the live event display thread. I made slight modifications in the live prediction to halt the continuous beep thread, which was causing the real-time event display problem at that time."
4.Pending task: Issue with repeat training mode not solved. Problem: Module beeps continuously after it stops (pending task)-[Status-Inprogress]

16-11-2023
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Now, for training and live event prediction, the audio is not playing due to an issue with the play audio module. I am currently fixing these issues
2."Discussions and planning with PL and teammates are ongoing regarding the integration of Deep Learning modules into the MindMe software."
3."Preparation for an EEG data viewer module code walkthrough is underway. This session aims to explain how the module works, its design, and implementation to team members."

17-11-2023
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1."Preparation for an EEG data viewer module code walkthrough is underway. This session aims to explain how the module works, its design, and implementation to team members."
2.Pending task: Issue with repeat training mode not solved. Problem: Module beeps continuously after it stops (pending task)-[Status-Inprogress]
3.Now working on implementing a delay between the LSL live EEG stream and the beep.
4.Explore and study how to implement delay in live EEG data processing in MindMe Software
5.Solved issue in importing other file in EEG data viewer during calling it into main.py
20-11-2023
-----------------------------------------------------------------------------------------------------------------------------------------
1.Resolved the issue with the repeat training mode.
2.Given a code walkthrough for the team members.(EEG Data Viewer module-about how the data browsing ,reading and plotting various format EEG data )

21-11-2023
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Team meeting with PL to discuss integrating issues of the Deep Learning module into MindME. The main problem is that we are providing raw EEG data for training without feature extraction, which is causing an issue. PL suggested modifying the execute pipeline by adding an if-else statement, and we discussed what parameters need to be changed in the execute pipeline to accommodate the Deep Learning module. PL assigned me to move the eye close mapping code into a separate file and import and use it from there. The reason he gave was to reduce the number of lines in the execute module. I created a new .py file called eyes_open_close_mapping and placed the code inside it, and then imported it into the eyes closing and open mapping portion.
2.Resolved an issue in the live prediction module. Problem: The event was being predicted continuously without any delay, causing it to be out of sync with the training beep in the background. Solution: Implemented the time.sleep module to introduce a delay between predictions, bringing them in sync with the training beep.
3.The Real-Time UI fails to close automatically when the user presses the "Stop Live Prediction" button. The training, processing, and LSL data receiving threads are successfully stopped, but the RealTime UI remains visible and must be closed manually. This issue is being addressed to enable automatic closing of the RealTime UI module.

22-11-2023
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Resolved the issue where a JSON loading error occurred in the wavelet feature extraction module when wavelet features were selected for live prediction.
2.Trained a deep learning model using LSTM and data provided by volunteers(Volunteer 2,session 1 and Volunteer 5 session3), and then evaluated its performance.
3.The Real-Time UI fails to close automatically when the user presses the "Stop Live Prediction" button. The training, processing, and LSL data receiving threads are successfully stopped, but the RealTime UI remains visible and must be closed manually. This issue is being addressed to enable automatic closing of the RealTime UI module.

23-11-2023
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Trained government medical college Neuro EEG Lab Technicians on using various tools in MindMe software. Additionally, installed PyCharm and successfully executed MindMe software on their laptops. Resolved package installation issues encountered on their laptops.
2.The EEG Lab Technician recorded my EEG data using OpenBCI software. Data was recorded for both voice-absent and voice-present conditions and then exported.
3.The Real-Time UI fails to close automatically when the user presses the "Stop Live Prediction" button. The training, processing, and LSL data receiving threads are successfully stopped, but the RealTime UI remains visible and must be closed manually. This issue is being addressed to enable automatic closing of the RealTime UI module.

28-11-2023
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.The Real-Time UI fails to close automatically when the user presses the "Stop Live Prediction" button. The training, processing, and LSL data receiving threads are successfully stopped, but the RealTime UI remains visible and must be closed manually. This issue is being addressed to enable automatic closing of the RealTime UI module.(pending task-parallely resolving it)
2.Two new modifications have been implemented. Firstly, if the user selects to perform feature extraction only, the program will extract features without performing any additional processing. Secondly, if any file is empty, the program will automatically display the name of the empty file in the console.
3.Preparing DDD for Wavelet Module,channel selection Module

29-11-2023
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1."Preparing DDD for sub-modules: channel selection, preprocessing, wavelet processing, up/down sampling, hardware selection, and creating train and test sets (splitting image data, creating single subject data, and creating data for all subjects)."
2."Resolving the issue of empty folders occurring in the feature extraction module."
3.Identify imbalanced epoch files during the file reading period in Execute pipeline(New task)

*30-11-2023
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Implemented imbalance epoch file identification code in execute_pipeline.py, execute_ppl() file reading portion.

2.Preparing DDD for EEG Data viewer module. DDD completed functionalities: load EEG data, restore original data, exit, plot, plot PSD, plot FFT, preprocessing, Notch filter, low-pass filter, high-pass filter.

3.Preparing document on signal processing techniques used in Mine Me software*

02-01-2023
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Today's progress involved the integration of eye blink detection with a real-time EEG prediction module within the MindMe software. Additionally, I created a live prediction method selection UI and developed its corresponding backend code
2.Threshold based eyeblink detection module integrated with real-time eeg data receiving thread function

Daily Report - January 3, 2024
------------------------------------------------------------------------------------------------------------------

1. Worked on Resolving Thread Issue in Live Threshold-Based Eye Blink Detection Module:
Explored various methods to address a thread issue encountered in the live threshold-based eye blink detection module. Conducted research on Stack Overflow, GitHub, and other resources to understand the root cause. Made progress in investigating the issue and partially resolved it. Efforts were focused on ensuring smoother execution of threads for improved functionality. Ongoing work is being done to find a comprehensive solution.

2. Bug Fix Reported by Rashmi in Filter UI:
Addressed an issue reported by Rashmi related to the filter UI. Successfully resolved the checkbox issue in the filter module, ensuring proper functionality.

3. Code Modification in PeakFindLivePrediction.py and Machine Pipeline Module:
Implemented modifications in the PeakFindLivePrediction.py module and the machine pipeline module for better performance and enhanced functionality.

Daily Report - January 4, 2024
-------------------------------------------------------------------------------------------------------------

1. Resolved Thread Issue in Live Threshold-Based Eye Blink Detection Module:
Successfully addressed a thread issue in the live threshold-based eye blink detection module. The issue was resolved by eliminating the usage of the Matplotlib algorithm from the code. Additionally, integrated the garbage collection (gc) module into the threshold-based eye blink detection method for optimization and improved memory management.
2. Initiating Coding for Channel Selection in Threshold-Based Eye Blink Detection:
Commenced coding to enable channel selection functionality for the threshold-based eye blink detection module. Started working on accommodating various combinations such as individual channel selection (e.g., 1, 2) and range-based channel selection (e.g., 1-2). This feature enhancement aims to provide users the capability to select specific channels, facilitating a more customized eye blink detection process.

Daily Report - January 8, 2024
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Solved an issue in the preprocessing module during EyeBlinkDetectionOfflineUI.py when selecting a specific channel and applying a filter.
2.Implemented a specific channel selection feature in the real-time EEG analysis module (machine_pipeline.py).
3.Integrated the preprocessing module with the PeakFindLivePrediction function using a threshold-based approach
4.The try-except block has been removed from the Eye Blink PeakFind() function, and the error-causing expect block has been resolved using an if-else condition

Daily Report - January 9, 2024
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Addressing the eye blink list appending issue in MachinePipeline.py during real-time EEG analysis.
2.Successfully completed the integration of the EyeBlinkPostprocessing.py module with MachinePipeline.py for enhanced functionality in the context of selecting dominants eye blinks in real-time stream

Daily Report - January 10, 2024
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Solved eye blink list appending issue in MachinePipeline.py
2.Solved hashable list issue in MachinePipeline.py module
3.Solved plotting issue in offline PeakFind.py module
4.Solved count issue in MachinePipeline.py module
5.Integrated RealTimeUI.py module with live threshold based eyeblink detection module
6.Presented the algorithm and functionality of the live threshold-based eyeblink detection system to PL. PL suggested some crucial changes, including implementing a beep in the process thread to ensure proper functionality. Additionally, recommendations were made to display the number of eyeblinks in the real-time UI and to disable the beeping thread functionality from the MachinePipeline.py module along with the Volunteer training option in the hardware settings module.

Daily Report - January 16, 2024
-----------------------------------------------------------------------------------------------------------------------------------------
1.Explored and implemented the timer module in Python, integrating it into the machine pipeline module.
2.Resolved an issue in the decision window of the machine pipeline by replacing the count mechanism with the timer module.

Daily Report - January 17, 2024
-----------------------------------------------------------------------------------------------------------------------------------
1.Successfully validated the live stream for eyeblink detection.
2.Explored various types of buffers in Python and seamlessly implemented io.BytesIO buffer in the machine pipeline module.
3.Fixed a problem reported by Dhanya in the feature extraction part.
4.Currently working on resolving the secondary buffer issue in the machine pipeline module.

Daily Report - January 18, 2024
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------
1. Resolved the secondary buffer issue in the machine pipeline module.
- Investigated and fixed issues related to the secondary buffer in the machine pipeline, ensuring it functions as expected.

2. Explored stream clearing and appending functionality.
- Assigned task by PL to investigate and understand how stream clearing and appending operations work, ensuring familiarity with the relevant concepts.

3. Addressed an issue reported by Reshmi regarding the secondary buffer not providing the latest stream data.
- Identified and fixed the reported problem by adding the `seek(0)` operation to the secondary buffer, ensuring it now provides the most up-to-date stream data.

Daily Report - January 30, 2024
-------------------------------------------------------------------------------------------------------------
1.Prepared the monthly report for team activities.
2.Integrated the image generation UI with Rashmi. After integrating the image generation UI into MindMe, properties such as alignment are not functioning properly when called from the MindMe software.
3.Prepared the CCI for the train set creation module.

Daily Report - January 31, 2024
-------------------------------------------------------------------------------------------------------------
1. Solved image generation issue in the image generation UI module,after integrating.
2. Prepared the CCI for the filtering module.

Daily Report - February 1, 2024
-------------------------------------------------------------------------------------------------------------
1.Testing and correcting realt-time eyeblink detection module in machine pipeline .
2.Solving alignment issue in image generation UI module.

Daily Report - February 2, 2024
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

1.Machine Pipeline Update Error:After updating the machine pipeline, an error started to appear in the update blink detected thread. The error message read: "TypeError: eBrainVoice.module.execute.machinePipeline.ExecuteRealTimeAnalysis.update_blink_detected() argument after * must be an iterable, not int."Upon investigation, I identified the root cause of this error. It was occurring because we were passing an integer value for "blink detected" instead of a string to the thread within the update blink detected function.To resolve this issue, I made the necessary changes to ensure that the "blink detected" parameter is passed as a string to the thread. After implementing these adjustments, the error mentioned above was successfully resolved.

Daily Report - February 5, 2024
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Resolved issue with offline eyeblink module: Rectified the problem where eye blink plots were not displaying due to the warning "UserWarning: Matplotlib is currently using agg, which is a non-GUI backend, so cannot show the figure." This issue was addressed by setting a suitable backend like TkAgg.

2.Knowledge Transfer Session: Participated in a comprehensive knowledge transfer session facilitated by Rashmi regarding eye blink detection and data recording protocols.

Daily Report - February 8, 2024
------------------------------------------------------
Task Progress:-
-------------------------------------------------------
1.Currently implementing code in the machine pipeline to skip the first 5 seconds of LSL stream data. This adjustment aims to enhance data processing efficiency by excluding initial non-essential data.

2.Rashmi provided detailed explanations on the following modules:

• channelwiseSpectrogram.py: Explored the functionalities and implementation details of the module.
• multichannel_scalogram.py: Understood the purpose and usage of the multichannel scalogram module.
• epoch_to_monochrome_image.py: Discussed the functionality and potential applications of the epoch to monochrome image conversion module.
• epochs_To_scalpMaps.py: Explored the functionalities and use cases of the epochs to scalp maps module.
1. epochs_To_grayscale_image.py: Understood the functionality and integration aspects of the epochs to grayscale image module.

Daily Report - February 9, 2024
----------------------------------------------------------------------------------------------------------------------------
1.Rashmi explained epochs_To_scalpMaps.py module: Explored the functionalities and implementation details of the module.
2.Implemented and tested in live, threshold-based eyeblink detection, with the removal of the first 5 seconds of data from self.processeddata. The plot will display the initial data, but it won't be processed.

Daily Report - February 12, 2024
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Implementing Markers in the live eyeBlink stream for improved tracking and analysis after deciding to focus solely on labeling eyeblinks, following PL's suggestion to drop the task due to previous unsuccessful attempts.Unfortunately, attempts to mark(Red line vertically on eyeblink peak) live eyeblinks were unsuccessful. Therefore, the decision was made to focus solely on labeling eyeblinks for analysis and tracking purposes.
2.Conducted testing of the live eye blink detection algorithm, utilizing Archana as the subject for data collection and evaluation.
3.In the live eyeBlink detection, the removal of 2 seconds of initial artifact data is causing an issue as it consistently eliminates 2 seconds of data each time. I am currently working on resolving this issue.

Daily Report - February 13, 2024
----------------------------------------------------------------------------------------------------------------------------------
1.Resolved bug in live eyeBlink detection: Identified and addressed the issue where the removal of 2 seconds of initial artifact data consistently eliminated 2 seconds of data in live eyeBlink detection. Currently working on implementing a fix to resolve this issue.
2.Conducted testing of the live eye blink detection algorithm, utilizing Archana and Reshmi as the subject for data collection and evaluation.

Daily Report - February 14, 2024
----------------------------------------------------------------------------------------------------------------------------
1.PL assigned me to explore and study different methods for removing involuntary blinks from live stream
2.Conducted testing of the live eye blink detection algorithm, utilizing Aswin,Anusree,Kausalya and Gouri as the subject for data collection and evaluation.
3.Reshmi Explained involuntary eyeblink removal from offline data using MATLAB approach explained

Daily Report - February 15, 2024
-----------------------------------------------------------------------
1.Attended a weekly meeting with PL to discuss project components we can demonstrate at the March 1st conference, as well as algorithm fine-tuning steps.
2.Attended the montage selection session for inner speech and eye blink by the EEG technician.
3.Explore independent component analysis

Daily Report - February 16, 2024
-----------------------------------------------------------------------
1.Explored and studied independent component analysis (implementation pending)
2.PL assigned me to implement a beep in live eyeblink detection
3.Discussion with PL regarding the preparation of a Call for Project Proposals for Innovative R&D in emerging areas of Information Technology (IT).
4.Preparation of Part A (1-14 pages) and milestones for the Centre for Brain-Computer Interface Research project proposal.

Daily Report - February 19, 2024
----------------------------------------------------------------------------------------------
1.Tested live prediction of inner speech using the OpenBCI device. Encountered an issue during the testing phase, which was promptly reported to the project lead (PL).

2.Updated the proposal for the Centre for Brain-Computer Research, focusing exclusively on the organization details, chief investigators, nature of the project, project objectives, and biographies of the chief investigators and co-investigators.

3.Prepared a document outlining the roles and responsibilities of each team member for the technical talk scheduled for March 1st.

Daily Report - February 20, 2024
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Resolved an error encountered during live inner speech processing. Initially, a ValueError occurred, indicating that the input data had 20 features, while the RandomForestClassifier expected only 8 features. This issue was resolved by reducing the stream size.
2.Participated in an ISO meeting organized by the project lead (PL).
3.Conducted testing of the inner speech model using the random forest classifier, employing the live prediction method.
4.Added header in python files

Daily Report - February 21, 2024
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Implemented beep functionality in live stream in the machinepipeline.py module.
2.Received a link shared by PL regarding the inner speech BCI dataset. Assigned to plot the available dataset.

Daily Report - February 22, 202
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Received recommendations from PL to amend the live secondary buffer due to the current buffer's loss of streaming data.
2.Participated in a team meeting alongside FLA and SLA, focusing on the March 1st technical talk titled "Sensors and Instrumentation for Human Machine Interfaces" by Professor Dr. Suresh Devasahayam, Professor of Electrical Engineering at IIT-Jammu. During the meeting, PL assigned specific roles and responsibilities to each team member for the event.

Daily Report - February 23, 202
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1. Changed `self.secondary_buffer = io.BytesIO()` to `self.secondary_buffer = np.zeros((self.n_samples, self.n_chan))` and re-arranged secondary code to solve buffer clearing issue.

2. Attempted to implement beep in livestream but encountered failure. Work in progress.

3.Now i'm resolving stream shifting issue in live eye blink detection algorithm

4.Preparing EMG based drowsiness detection technical content for Center for BCI Proposal

Daily Report - February 26, 202
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Leave
Daily Report - February 27, 202
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Secondary Buffer Maintenance Implemented:
-Developed and implemented code to clear the secondary buffer, ensuring optimal data management and system performance.
2.Meeting with PL and Team Members:

-Attended a productive meeting with PL and team members to discuss the roles and responsibilities for the upcoming March 1st technical talk titled "Sensors and Instrumentation for Human Machine Interfaces" by Professor Dr. Suresh Devasahayam, Professor
of Electrical Engineering at IIT-Jammu.
-Discussed the agenda and expectations for the technical talk, clarifying individual contributions and coordination efforts among team members.
-Outlined key points and action items to ensure effective preparation and seamless execution of the event.
3.Task in Progress: Peak Find Python Function Exploration,
-Currently exploring peak find Python function as assigned by PL

Daily Report - February 28, 202
-----------------------------------------------------------------------------------------------------------------------------------------------------------
1.Explored the peak find module and applied various parameters to enhance the live eyeblink detection module.

2.Prepared the live eyeblink detection algorithm for the upcoming demo.

3.Assigned by PL to study and explain the following features for tomorrow's presentation:
-Mean Absolute Value (TD)
- Root Mean Square (TD)
- Squared Integral (TD)
- Average Amplitude Change (TD)
- Number of zero-crossings (TD)
- Lempel-Ziv complexity (TD)
-Petrosian fractal dimension (TD)
-Katz fractal dimension (TD)
-Power Spectral Density (FD)
-Spectral Entropy-Welch (FD)
-EegData_Analysis Entropy (FD)
- Peak Frequency (FD)
-Spectral Centroid (FD)
- Spectral Flatness (FD)
- Wavelet 25th percentile value (WD)
-Wavelet 75th percentile value (WD)

Daily Report - March 1, 2024
--------------------------------------------------------------------
1.Demonstrated Eye blink detection system to Dr. Suresh Devasahayam sir
2..Attended IEEE Conference ,Sensors and Instrumentation for Human Machine Interfaces" by Professor Dr. Suresh Devasahayam, Professor
of Electrical Engineering at IIT-Jammu.

Daily Report - March 04, 2024
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Attended the BCI project team meeting with SLA and FLA.
2.Participated in an online meeting with the North East Hospital, Tezpur Medical College Hospital, regarding the proposal for dementia screening. Discussion held in conjunction with SLA and FLA.
3.Assigned by PL to prepare an IEEE journal paper on "Real-Time EEG-Based Eye Blink Detection and Interpretation for Human-Computer Interaction and Communication." Prepared abstract and sent it to PL for approval.

Daily Report - March 05, 2024
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.PL's suggested modifications in the Abstract on Real-Time EEG-Based Eye Blink Detection.
2.Initiated exploration and study of IEEE transactions guidelines as per PL's assignment.

Daily Report - March 06, 2024
------------------------------------------------------------------------------------------------------------
1.Completed: Reviewed general writing guidelines for IEEE Transactions journals.
2.Revised: Edited Real-Time EEG-Based Eye Blink Detection and Interpretation for Human-Computer Interaction and Communication journal abstract.
3.Action: I applied the learned guidelines to the revised abstract.

Daily Report - March 07, 2024
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Prepare a short note about general writing guidelines for IEEE Transactions journals and share it with all team members. 

Daily Report - March 11, 2024
-----------------------------------------------------------------------------------------------------------------------------------------------
1. reviewed general writing guidelines for IEEE Transactions journals
2. Modified the Abstract on Real-Time EEG-Based Eye Blink Detection and sent it to PL for approval.

Daily Report - March 12, 2024
-----------------------------------------------------------------------------------------------------------------------------------------------
1.Presented the general writing guidelines for IEEE Transactions to the team members.
2.Prepared the introduction for the paper on real-time EEG-based eye blink detection.

Daily Report - March 13, 2024
-----------------------------------------------------------------------------------------------------------------------------------------------
1.Prepared the literature review of eyeblink detection for the paper on real-time EEG-based eye blink detection.
2.Attended group discussion with GH about design of CCET

Daily Report - March 14, 2024
-----------------------------------------------------------------------------------------------------------------------------------------------
1.Facilitated and organized technical arrangements including microphones and laptops for Ms. Khushboo Srivastava, CEO of One Learning Solutions, for her technical talk.
2.Finished compiling literature review on Machine Learning (ML) based eye blink detection from EEG signals for an IEEE journal publication. Currently progressing with the writing process.

Daily Report - March 19, 2024
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Reviewed and verified the budget proposal for project "2022-1849" titled "Indigenous Anesthesia Depth Monitoring System using Brain Monitoring Technology".

2.Completed the nomination form for the Training Programme to attend IFNRCON 2024, the 12th Annual Conference of the Indian Federation of Neurorehabilitation. The conference, themed "Reshaping Neurorehab through Technology," will be held at JIPMER, Pondicherry from April 12th to 14th, 2024. Obtained necessary approvals from SLA, FLA, and GH, and submitted the form to the finance department for further processing.

3.Compiled the literature review on Threshold-Based Eye Blink Detection(existing studies) from EEG Signals for an upcoming IEEE journal publication. Currently, in the writing phase, Currently progressing with the writing process.

Daily Report - March 20, 2024
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Conducted a thorough check on train availability and explored hotels in close proximity to JIPMER, Pondicherry, in preparation for attending the IFNRCON 2024. This event marks the 12th Annual Conference of the Indian Federation of Neurorehabilitation, scheduled with the theme "Reshaping Neurorehab through Technology," to be hosted at JIPMER from April 12th to 14th, 2024.

2. Methodology part of the "Real-Time EEG-Based Eye Blink Detection and Interpretation for Human-Computer Interaction and Communication" journal paper writing in-progress

Daily Report - March 21, 2024
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Methodology part of the "Real-Time EEG-Based Eye Blink Detection and Interpretation for Human-Computer Interaction and Communication" journal paper writing in-progress
-ML based approach completed
-Threshold based approach started

Daily Report - March 26, 2024
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Methodology part of the "Real-Time EEG-Based Eye Blink Detection and Interpretation for Human-Computer Interaction and Communication" journal paper writing in-progress

Daily Report - March 27, 2024
-------------------------------------------------------------------------------------------------
1. Followed up on account details and program schedule for the upcoming 12th Annual Conference. Prepared offline registration forms for 3 participants.
2. Generated scalogram images for all eye blink volunteer epoch data and split the data into train and test sets. Created train set images for training using CNN deep learning algorithm.
3. Incorporated feature extraction equations into the "Real-Time EEG-Based Eye Blink Detection and Interpretation for Human-Computer Interaction and Communication" journal paper.
4.March Monthly Task Report: Documented all team activities in a Word document.

Daily Report - March 28, 2024
-------------------------------------------------------------------------------------------------
1.Provide d new scalogram image data for testing using CNN deep learning algorithm.
2.Explored threshold-based eye blink journal papers for reference to obtain the threshold equation. Added explanations and equations for Gaussian filtering, normalization, detrending, and baseline correction preprocessing methods to the "Real-Time EEG-Based Eye Blink Detection and Interpretation for Human-Computer Interaction and Communication" journal paper
3.PL reported errors during the testing of the MindMe software in the feature extraction module. Specifically, runtime warnings are appearing in the console.

Daily Report - April 1, 2024
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.New scalogram image data for testing using the CNN deep learning algorithm has been provided. An issue occurred in the DL module during testing, resulting in an error. I reported this issue to Kausalya
2.Modified explanations and equations for Gaussian filtering, normalization, detrending, and baseline correction preprocessing methods are provided in the journal paper 'Real-Time EEG-Based Eye Blink Detection and Interpretation for Human-Computer Interaction and Communication

Daily Report - April 2, 2024
--------------------------------------------------------------------------------------------------------------------------------------------------------
1.PL assigned me to explore and study how to read and write ML model into pickle
2.Modified 'Real-Time EEG-Based Eye Blink Detection and Interpretation for Human-Computer Interaction and Communication

Daily Report - April 3, 2024
---------------------------------------------------------------------------------------------------------------------------------------
1. Explore and study methods to convert a tkinter standalone application into a web-based application.
2. Followed up on the payment status for the IFNRCON 2024 conference.
3. Revised and prepared the PowerPoint presentation for the IFNRCON 2024 conference.
4. Explored the peak find module in Python for developing equations for threshold detection in real-time EEG-Based Eye Blink Detection and Interpretation, intended for the Human-Computer Interaction and Communication journal paper.
5. Parallely transferring the HTG047D folder from the HTG share to the Seagate hard disk (3.3TB) for backup purposes.
6.Attended a signal processing class on Singular Value Decomposition, taught by Anusree V.
7.Tested the OpenBCI Cython board after attaching it to the SLA body cover.

Daily Report - April 04, 2024
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1. Modified the journal paper "Real-Time EEG-Based Eye Blink Detection and Interpretation for Brain-Computer Interaction and Communication" for submission to IEEE Transactions.
2. Prepared and submitted the PowerPoint presentation for the IFNRCON 2024 conference to the project lead for review.
3. Attended the filter design class taught by Gouri.
4.PL assigned me to teach the kurtosis feature tomorrow.

Daily Report - April 05, 2024
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1. Attended the basic filter design class taught by Gouri.(Morning section)
2. Assigned by PL to correct the visualization tool and prepare it for the IFNRCON 2024 conference.
3. Studying and modifying the Kurtosis Feature Extraction in EEG Data presentation.
4. Discussed with PL about the modifications required for the presentation prepared for the IFNRCON 2024 conference.
5.Preparing python modules ready for demo for the IFNRCON 2024 conference
6.Exploring and studying User Manual of Quantitative EEG and multi-model Neuro Imaging Biomarkers of Memory dysfunction in epilepsy
7.Completed and submitted for FLA and SLA review, the journal paper titled 'Real-Time EEG-Based Eye Blink Detection and Interpretation for Brain-Computer Interaction and Communication' is intended for submission to IEEE Transactions.
8.Testing and debugging of the real-time eye blink detection algorithm
9.Attended a signal processing class on Singular Value Decomposition, taught by Anusree V(Evening section)

Daily Report - April 17, 2024
----------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Solved preprocessing module issue reported by Kausalya.
2.Submitted for review of journal paper titled 'Real-Time EEG-Based Eye Blink Detection and Interpretation for Human-Computer Interaction and Communication' to FLA and SLA.
3.PL assigned me to explore and study producer-consumer stream in Python."

Daily Report - April 18, 2024
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1. Started design and implementation of MindMeLite software for testing EEG data offline. Created three Python files: Run_MindMeLite, FeatureExtraction, and TimeDomainFeatureExtraction modules. Currently resolving an issue in the feature extraction module related to reshaping.

2. Integrate MindMeLite with the machine learning module developed by Miss. Dhanya M R.

Daily Report - April 22, 2024
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Explored Queue in Python and attempted to implement it in the real-time eyeblink detection module. Encountered challenges and unresolved issues during implementation.
2.Investigated and worked on resolving issues with the real-time eyeblink detection module. Analyzed potential causes of the problem and devised strategies for troubleshooting and resolution.
3.Integrated MindMeLite with the machine learning module developed by Miss. Dhanya M R.

Daily Report May 01, 2024
----------------------------------------------------------
1.Explored Queue in Python and attempted to implement it in the real-time eyeblink detection module.

Daily Report May 02, 2024
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Implemented enqueue and dequeue functions for real-time eyeblink detection module.
2.Developed separate thread functions for the enqueue and dequeue processes.
3.Attempted to pass the queue to the real-time UI module for visualization purposes. However, encountered an error after 5 minutes indicating that the queue is full. Currently investigating and resolving this issue. Determining whether to clear the queue during enqueue or dequeue operations.

Daily Report May 08, 2024
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Successfully resolved the issue with OpenBCI data recording and conducted thorough testing.
2.Volunteered for the recording of subvocalized speech data, including voice absent data under both eye open and closed conditions, as well as voice present data under both eye open and closed conditions.
3.Made modifications to the machine learning section of the journal paper. Additionally, progressed on the mathematical aspect of the threshold-based method. Currently exploring mathematical equations suitable for eye blink detection from EEG data.

Daily Report May 09, 2024
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Research Progress on Random Forest Classifier Algorithm for Journal Paper:Today, I spent time studying the math behind the Random Forest classifier algorithm. This is important for my upcoming journal paper. I want to understand how the algorithm works so I can explain it well in my paper.
2.Additionally I explored and read IEEE journal papers related to threshold-based eye blink detection from EEG signals

Daily Report May 13, 2024
-----------------------------------------------------------------------------------------------------
Leave

Daily Report May 14, 2024
---------------------------------------------------------------------------------------------------
1.Compiled a thorough guide for the forthcoming ICCON BCI Conference 2024, providing essential details about tourist attractions in Thiruvananthapuram, to be uploaded on the ICCON BCI conference website.
2.Assigned by PL to create an installation guide for MindMe software, focusing on clear and user-friendly instructions to facilitate seamless implementation and usage for end-users
3.Modified machine learning section of the journal paper.
4.Restructured and formatted journal draft according to IEEE transaction guidelines

Daily Report May 15, 2024
---------------------------------------------------------------------------------------------------
1.Finished editing the journal paper entitled 'Real-Time EEG-Based Eye Blink Detection and Interpretation for Brain-Computer Interaction and Communication.'

2.Following PL's suggestion, aiming to expand the content to a total of 20 pages. Currently, the journal paper comprises 12 pages.

3.Incorporated Machine Learning (ML) algorithms and threshold algorithms into the journal paper. Additionally, researching threshold-based eye blink mathematical formulas to include in the paper.
4.PL assigned me to research a similar journal paper and delve into the analysis of results and statistics.

Daily Report June 24, 2024
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Sent Author consent form to GMCT for getting signatures from doctors; 90% of journal paper submission completed.
2.Completed the chart preparation for IEEE cross-check.
3.Queue coding and testing in progress.

Daily Report June 25, 2024
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Government Medical College doctors' addresses were incorrect, so they were modified and sent to the GMCT team for signatures on the Author consent form. Dr. Praveen Panicker is on leave, so this task is pending.
2.Decided to use a double-ended queue (deque or input-output queue) in the machine pipeline module(https://www.geeksforgeeks.org/deque-set-1-introduction-applications/). Code design and development are in progress.
3.OpenBCI data recorded as per CDAC-Delhi's request for two 5-minute sessions. During that time, OpenBCI did not work, so the issue was resolved with the help of team members Anusree (recorded her data) and Aswin (helped solve the hardware issue).
4.Kausalya shared the 3rd PRSG presentation and explained it according to the PL's suggestions.
5.Gathered the team's tasks completed this month and compiled the monthly report, forwarding it to the project leader for review.

Daily Report June 26, 2024
------------------------------------------------------------------------------------------------------------------------------------------------
1.Finalized the design and development of the double-ended queue in the machine pipeline module with the PL.
2.Implemented event selection functionality in the test module.
3.Set up the Acxxel PC with Python and PyCharm.
4.Helped Dhanya to resolve an error in the time domain feature extraction module. The error was: 'AttributeError: Scipy has no attribute sum.

Daily Report June 26, 2024
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.As part of the knowledge-sharing session, I presented on the topics of types of signals, characteristics of signals, operations on signals, and classification of signals.
2.Attended sessions on Frequency Domain Signal Analysis conducted by Miss Anusree, Non-Linear Signals conducted by Miss Dhanya MR, Communication Protocols conducted by Miss Gouri, and Analog to Digital Converter conducted by Mr. Ashwin Sharma.
3.Solved the error in the event selection module and tested it on Miss Kausalya's and Miss Dhanya's PCs. It is working fine. Now, testing and review from the Project Leader (PL) are pending.
4.IEEE journal paper submission is pending.
5.Queue code development is pending.

Daily Report June 27, 2024
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.As part of the knowledge-sharing session, I presented on the topics of types of signals, characteristics of signals, operations on signals, and classification of signals.
2.Attended sessions on Frequency Domain Signal Analysis conducted by Miss Anusree, Non-Linear Signals conducted by Miss Dhanya MR, Communication Protocols conducted by Miss Gouri, and Analog to Digital Converter conducted by Mr. Ashwin Sharma.
3.Solved the error in the event selection module and tested it on Miss Kausalya's and Miss Dhanya's PCs. It is working fine.
4.IEEE journal paper submission is pending.
5.Queue code development is pending.

Daily Report June 28, 2024
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Designed and developed a method to exclude trained files from the test set in the test module as assigned by the PL.
2.Integrated the deep learning module with the real-time data stream module as assigned by the PL.
3.Designed and developed a queue in the real-time eye blink thread module as assigned by the PL.
4.Designed and developed the simulator module as assigned by the PL.
5.Resolved a bug reported by the PL in the event testing module (Error: accuracy = str((gl.right_prediction / total) * 100) ZeroDivisionError: division by zero).
6.Identified a mistake in the Vellam event folder names, informed the GMCT EEG technicians, and corrected the folders in the HTG shared data repository.
7.Design and development of a method for reading data from the Brain Product device in Python.
8.Updated the MiMe software user manual.
9.Take steps to set up the server-capable PC, with assistance from Rajesh Sir from the network group.

Daily Report June 29, 2024
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Resolved a bug reported by the PL in the event testing module (Error: accuracy = str((gl.right_prediction / total) * 100) ZeroDivisionError: division by zero).-Completed Task
2.Identified a mistake in the Vellam event folder names, informed the GMCT EEG technicians, and corrected the folders in the HTG shared data repository.-Completed Task
3.Updated the MiMe software user manual. Completed Task
4.Take steps to set up the server-capable PC, with assistance from Rajesh Sir from the network group. - Completed Task
5.Design and development of a method to exclude trained files from the test set in the test module as assigned by the PL-Pending Task
6.Integrate the deep learning module with the real-time data stream module as assigned by the PL- Pending Task
7.Design and development of the simulator module as assigned by the PL.-Pending Task
8.Design and development of a queue in the real-time eye blink thread module as assigned by the PL.- Pending Task
9.Design and development of a method for reading data from the Brain Product device in Python. -Pending Task

Daily Report July 01, 2024
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Designed and developed a method to exclude trained files from the test set in the test module. During the process, an error occurred in the test module (UnboundLocalError: local variable 'test_json' referenced before assignment). This error has been resolved. Currently, the listing of trained files portion is pending (85% completed).
2.Attended a Volterra series knowledge sharing session conducted by Ms. Dhanya MR.
3.Forwarded the author consent form to team members for signing.

Daily Report: July 02, 2024
-------------------------------------------------------------------------------------------
1.Integrated the UI module of deep learning into the real-time module and completed raw signal live testing with various deep learning algorithms.
2.Testing of feature-extracted data with the deep learning module in real-time is pending(algorithm development is pending).
3.The image-based approach in real-time integration with the deep learning module is also pending.

Daily Report: July 03, 2024
--------------------------------------------------------------------------------------------------------------------------------
1.Integrated feature extraction-based deep learning module into the real-time module (machine_pipeline.py).
2.The image-based approach in real-time integration with the deep learning module is in progress.
3.Currently solving an error in the update_ui_label thread related to the playsound library.{Error:Exception in thread Thread-36 (update_ui_label): playsound.PlaysoundException:
Error 277 for command:open "rightTurn_audio.mp3" alias playsound_0.09520148617161073 ,A problem occurred in initializing MCI.}

Daily Report: July 04, 2024
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.PL assigned me to understand the image-based deep learning algorithm workflow in offline mode in Mime software and document the mandatory functionalities needed for integrating the image-based module with the real-time module. This task was transferred from Kausalya. Progress is 70% complete, with a prepared word document and ongoing implementation.

2.Morning discussion with PL about the integration of the simulator module and real-time EEG signal data streaming. We discussed steps for the design and development of this module. The task has not started yet; only the design phase is completed.

3.Designed and developed offline reading and visualization of '.vmrk' and '.vhdr' format data from the Brain Product device. The task is completed, and testing of the implemented code is in progress.

4.Discussion with PL about the design and development steps of trainDataFilter.py. This module checks if trained data is present in the testing data. It will only allow new data through, blocking any trained data found in the testing data and prompting for new data.Development in progress

Daily Report: July 05, 2024
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Completed testing of the developed offline reading and visualization of '.vmrk' and '.vhdr' format data from the Brain Product device. (100%)-(Solved the error in the bandpass filtering module and also fixed it)
2.Transferred development of trainDataFilter.py module task to Miss Dhanya.
3.Started the design and development of the eye blink-based simulator module.
4.Submitted IEEE journal paper titled "EEG-Based Eye Blink Detection and Interpretation for Human-Computer Interaction and Communication" to the IEEE Journal of Biomedical and Health Informatics.
5.Completed channel-wise selection of Brain Product device in the real-time module.

Daily Report: July 08, 2024
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Integrated and modified simulator module, solved the issue: ConnectionRefusedError: [WinError 10061] No connection could be made because the target machine actively refused it. Integration of real-time module with simulator is in progress.

2.Shared details regarding EEG machines used for data acquisition with CDAC-Delhi team member.

3.Discussed 3rd PRSG activities with PL.

4.Checked status of manuscript titled, "EEG-Based Eye Blink Detection and Interpretation for Human-Computer Interaction and Communication", for possible publication in Journal of Biomedical and Health Informatics.

Daily Report: July 17, 2024
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Discussed and updated the current status of the simulator module. PL suggested changing the current file sharing method to streaming data.
2.PL assigned exploration of how to integrate the LSL networking data sharing method in the EEG simulator and started exploring and understanding the NeuroDecode code (available at https://pypi.org/project/neurodecode/).
3.Explored the IEEE Transactions on Cognitive and Developmental Systems guidelines for research paper submission.
4.Prepared a biography for Dr. Thomas Iype for inclusion in the "Electroencephalography Network-based Brain-Computer Interface for Inner Speech Identification" journal paper.
5.Started modifying the "Network-based Brain-Computer Interface for Inner Speech Identification" journal paper.
6.Received a revision of the manuscript titled "EEG-Based Eye Blink Detection and Interpretation for Human-Computer Interaction and Communication" for possible publication in the Journal of Biomedical and Health Informatics. Started modifying this journal paper.
7.Assigned by PL to explore common average filtering.

Daily Report: July 19, 2024
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Discussion with PL Regarding Simulator Module Development:Discussed the simulator module development with the Project Lead (PL). Based on the discussion, it was suggested to move the simulator module to a separate project file for more efficient data streaming. Planned works in the simulator module have started, and the task is currently in progress. As part of this work, we decided to stream data in chunks. The PL advised exploring and implementing the conversion of file data into chunks for streaming.

2.Developed Separate Python Project File for Simulator Module :Created an independent project file dedicated to the simulator module to enhance modularity and ease of maintenance.

3.Modified Simulator Data Sending Method : Implemented EEG data streaming through the LSL (Lab Streaming Layer) module, updating the simulator module UI accordingly. Added LSL and TCP/IP network label frames, along with a newly developed LSL network frame. The LSL network label frame now includes fields for stream ID, stream type, stream name, stream format, and channels.

Daily Report: July 25, 2024
---------------------------------------------------------------------------------------------------------------------------------------------------
1.Added a radio button to the ebrain_voice_pipeline.py in MiMe software for seizure detection.
2.Modified the downsampling and upsampling modules to handle datasets with multiple sampling rates (250 Hz, 256 Hz, 512 Hz, and 4096 Hz). The updated code allows for selecting the appropriate sampling rate data files for upsampling or downsampling 3.EEG data. This functionality is developed for creating all epoch file models and analysis purposes.
4.Currently resolving the sender-side LSL data streaming issue, exploring how to push streams as chunks in the outlet.

Daily Report: July 26, 2024
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Explored and implemented stream pushing as chunks in the outlet of LSL.
2.Developed visualization of .txt and .edf format EEG (epoch) data in the simulator (LSL-based EEG simulator). Faced and resolved a scaling issue in the .edf file visualization part.
3.Currently implementing a notch filter in stream visualization to address power line noise present in the .edf format EEG data.
4.PL identified and reported an issue with the time domain feature extraction in the energy feature, resolved it, and then committed and updated the code in SVN.
5.Prepared BMI team monthly report and forwarded to PL for review

Daily Report: July 29, 2024
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Explored Thread Function Implementations: Investigated various thread function implementations and communication techniques for integrating these logics into the EEG simulator module.

2.Issue Resolution in EEG Simulator: Addressing a problem where the EEG data values retrieved from the stream inlet are excessively high during EDF data reading. Currently working on normalizing the data or introducing a scale factor in the stream inlet to correct this issue.

3.Discussion with PL: Had a discussion with PL regarding the status of the EEG simulator module development. Provided an explanation of the working flow of the background threads involved in the module.

Daily Report: July 30, 2024
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Discussion with PL: Had a discussion with the Project Lead (PL) regarding the status of the EEG simulator module development, the circular import issue, and the development of producer-consumer based thread functions.

2.Current Implementation: The implementation of producer-consumer based thread functions in the simulator module is in progress. The shared queue module and producer thread function for inletting stream data and putting it in the shared queue have been implemented. Additionally, the consumer thread functions for visualization and analysis have been developed. Currently, I am resolving issues with the consumer thread for visualization.

Daily Report: July 31, 2024
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Morning meeting with PM and PL about the current work status.
2.Current Implementation: Resolved the issues with the consumer thread for visualization. Plotted .edf and .asc files in the EEG data viewer . However, when dynamically trying to change the window size, encountered the error ValueError: shape mismatch: objects cannot be broadcast to a single shape. Mismatch is between arg 0 with shape (640,) and arg 1 with shape (768,).

• # egg_channel_selection.py
• # eye_blink_based_threshold_detection.py
• # real_time_event_detection_ml_based.py
• # # feature_based_dl_test.py
  4.Currently working on:Solving bugs in the consumer_thread_for_analysis.py module.Preparing for tomorrow's demo to the PM, PL, and team members.
  5.Currently an issue in Real-time feature extraction module getting Nan values in feature extracted data frame, like below
  C:\Users\DELL\AppData\Local\Programs\Python\Python39\lib\site-packages\numpy\core\fromnumeric.py:3432: RuntimeWarning: Mean of empty slice.
  return _methods._mean(a, axis=axis, dtype=dtype,
  C:\Users\DELL\AppData\Local\Programs\Python\Python39\lib\site-packages\numpy\core\_methods.py:190: RuntimeWarning: invalid value encountered in double_scalars
  ret = ret.dtype.type(ret / rcount)
  df_all_feat_single_file_merged, 0 1 2 3
  0 -0.000017 NaN NaN NaN
  1 -0.000106 NaN NaN NaN
  2 -0.000069 NaN NaN NaN
  3 -0.000009 NaN NaN NaN

Daily Report: August 02, 2024
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Solved issue in real-time feature extraction module where NaN values were present in the feature extracted data frame.
2.Encountered another issue during real-time data testing: ValueError: X has 4 features, but SVC is expecting 8 features as input. Reported the issue to the project lead (PL) and discussed potential solutions. The PL suggested a better method to resolve the issue, which successfully solved the problem.
3.Implemented two background daemon threads in consumer_thread_for_analysis.py module for displaying dominating events in the EEG data viewer but event label not updating in EEG data viewer- task in progress.
4.Explored how to call a background thread from the main thread in Python for implementation in the real-time analysis module

Daily Report: August 05, 2024
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Task Completed: Implemented two background daemon threads in consumer_thread_for_analysis.py for displaying dominating events in the EEG data viewer, but the event label is not updating in the viewer.
2.Training Attended: Attended a gradeu web development Python training session taught by Miss Vaishnavi.
3.Error Resolved: Solved a real-time error reported by Miss Anusree in the SSVEP-based real-time EEG data analysis module.
4.Issue Resolved: Fixed a buffer issue in the eye blink detection module. Stream data initialization time was filled with zeros; this buffer is now correctly used for real-time analysis, ensuring all chunk data is appended and zeros are replaced.
5.Current Task: Working on resolving an issue in the real-time eye blink detection module where eye blinks are not being detected.(inprogress)
6.Task in Progress: PL reported an issue in feature extraction using new epoch data created by GMCT EEG technicians. Currently cross-checking data and debugging the feature extraction module.

Daily Report: August 06, 2024
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Conducted a demo of the simulator module, explaining its operation and parameters to the Project Lead (PL) and team members.
2.Encountered an issue in the EEG simulated data analysis portion where the decision window did not function as required. Discussed the issue with PL, who suggested an effective solution. Implemented the recommended logic, resulting in proper functioning
of the continuity check of the event list and the dominated event identification block of code.
3.Tested the ML and DL simulator algorithms with and without features after the above changes.In DL testing side changes needed,corrected that portions
4.Issue Resolved: Fixed a buffer issue in real-time EEG visualization. The stream data initialization time was previously filled with zeros; the buffer is now correctly used for real-time visualization, ensuring all chunk data is appended and zeros are replaced.
5.Prepared a justification report for an uncleared returnable gate pass and forwarded it to the PL for review.
6.Task in Progress: Working on resolving an issue in the real-time eye blink detection module where eye blinks are not being detected.-Preparing simulator and real-time module for tomorrow's demo

Daily Report: August 19, 2024
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Resolved the issue with clearing the label frame in the AI PLATFORM UI module.
2.Implemented the "Load Config" functionality in the AI PLATFORM UI module and updated gl.settings_json with new parameters from the UI, including the test input data file path and model path.
3.Explained the required functionalities for live or simulator data configuration to Anusree.

Daily Report: August 20, 2024
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Integrated the PyQt5-based UI developed by Miss Anusree with real-time modules including machinepipeline.py, feature.py (Time Domain, Frequency Domain, and Wavelet Domain modules pending), and threading.py. The GUI now receives streams from LSL, detects EEG events from the EEG stream, and displays the predicted events with audio feedback.

Daily Report: August 21, 2024
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Dhanya reported an issue with the test data post-processing button not clearing the frame correctly. The frame clearing issue was resolved.

2.In the PyQt5-based real-time test MiME UI, a new feature was added in the post-processing section. If "Text" is selected, it will display the event test; if "Audio" is selected, it will produce an event sound as feedback. If both options are selected, it will display the event test and generate the corresponding audio. The implementation logic for these functionalities was explained to Miss Anusree, and guidance was provided for the implementation process.

3.The Project Lead (PL) reported an issue in the EEG data viewer module where multiple parameter UIs were popping up during data selection. This issue was resolved, and the updated code was committed to SVN.

4.The Project Lead (PL) assigned the task of displaying seizure data visualization, with each stage represented by different color codes in the seizure data.

5.Seizure JSON dumping and loading work has been completed. Wavelet JSON dumping and loading are also completed, and testing is currently in progress.

Daily Report: August 28, 2024
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Provided the current work status to the FLA in the morning. Demonstrated the current stage of the software and reported existing bugs.

2.Prepared the project monthly report and forwarded it to the FLA for review.

3.Discussed AI platform code integration with the FLA.

4.Encountered an issue where scrolling horizontally through the Matplotlib built-in scroll bar caused problems with epoch creation. The FLA suggested implementing 'Previous' and 'Next' buttons for navigation, with mouse clicks restricted to within the EEG plot area, ignoring clicks outside the plot.Bug resolved.

5.FLA assigned me the task of implementing a window size entry feature in the EEG plot, similar to the MATLAB version. This task is currently in progress, and exploring how to increase the width and height of the matplotlib fig

Daily Report: August 29, 2024
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

1.Epoch Creation: Completed the epoch creation functionality in the python based EEG data viewer module.
2.Demonstration: Explained and demonstrated the epoch creation functionality to the FLA and team members before integrating filters and other functionalities
3.Visualization and Filtering: Integrated all filters into the updated visualization functionality. Noted that applying filters to the data currently takes about one minute for filtration,fft and psd functionalities not integrated(inprogress)

Daily Report: September 20, 2024
--------------------------------------------------------------------------------------------------------
MiMe Version 3 Progress:
------------------------------------------
1.Deep Learning Integration: Deep learning functionalities have been successfully integrated into MiMe Version 3.
2.Error Resolution: Debugged the classifyDL module to identify the root cause of the "Error:raise InvalidIndexError(self._requires_unique_msg) pandas.errors.InvalidIndexError: Reindexing 3only valid with uniquely valued Index objects." issue. The issue was traced back to inconsistencies in EEG data channels. Reported the error and solution proposal (implementing channel selection functionality) to the Project Lead (PL). The channel selection functionality was successfully implemented and tested in both deep learning testing and training modules, resolving the issue in MiMe Version 3.  
3.Seizure Bar Chart Creation: Implemented and explained the use of threads and a specific condition for seizure bar chart creation during testing in MiMe Version 3. Provided explanation to Leakshmi.
4.Knowledge Transfer: Explained and trained Leakshmi and Gowrisree on the background workflows of both EbrainVoice_pipline.py (specifically the activate_execute_btn_function) and execute.py modules.
5.Bug Fix and Training: Fixed a bug in the EEG Data viewer. Trained Gowrisree on epoch creation.

Daily Report: September 24, 2024
--------------------------------------------------------------------------------------------------------
1.Channel selection implemented in the deep learning modules both in the training and testing
2.Image loading in deep learning modules both in the training and testing integration and testing completed.

Daily Report: September 25, 2024
--------------------------------------------------------------------------------------------------------
1.Excluded UI functionality from the test module connected to the execution pipeline – task completed.
2.Implemented event-wise analysis functionality in the test module and connected it with the execution pipeline – task completed.
3.Integrated feature extraction functionality into the execution pipeline – task completed.
4.Integrated data balancing functionality with the execution pipeline – task completed.

Daily Report: September 30, 2024
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Identified and resolved UI issues related to seizure data image generation within the eye blink image generation module. The resolution applied to the following functionalities:epochs_to_channelise_scalogram,epochs_to_color_image,epochs_to_grayscale_image,epochs_to_monochrome_image,
epochs_to_scalpMaps,epochs_to_Spectrogram,multichannel_scalogram

Daily Report: October 01, 2024
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Integrated the offline threshold-based eye blink module and its functionalities into MiME Version 3.
2.Resolved an issue in EbrainVoicePipeline.py during the deep learning process, as reported by the Project Lead (PL).
3.Integration of the live module with updated settings configuration file is task in progress.
4.Explore and implement live Video Generation from EEG data with Python.

Daily Report: October 15, 2024
-----------------------------------------------------------------------------------------------------
1.Error correction in Mime version 3

Daily Report: October 21, 2024
-----------------------------------------------------------------------------------------------------
1.Assisted in resolving threading issues for progress bar generation in seizure detection.
2.Explored the use of the YOLO algorithm for EEG signal analysis.
3.Implemented code to receive live EEG data from LSL and convert it into color images.
4.Investigating additional signal processing methods for detecting eye blinks from EEG signals.
5.Discussed YOLO-based eye blink detection algorithms with Mr.Sambhu, and provided input EEG color image.

Daily Report: October 22, 2024
-------------------------------------------------------------------------------------------------------------------------------
1.Encountered an issue with YOLO package installation, which was resolved with the help of Mr. Sambhu.
2.Integrated YOLO model testing code with real-time LSL data to test eye blink detection from the stream (simulator). Created a YOLO model with limited data for testing, but the model did not detect eye blinks from the stream. To address this, generated more color EEG images and submitted them for further YOLO training.
3.While generating images, identified noise in the color EEG images. The preprocessing module was not yet implemented in the image generation UI. Currently working on integrating the preprocessing module into the image generation process.
4.Guided juniors in resolving their Python code issues, including reading and plotting anesthesia EEG data and implementing a seizure detection progress bar for both offline and real-time data, as per the instructions from PL.

Daily Report: October 22, 2024
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Integrated the preprocessing module into the image generation UI, focusing on the color image generation portion.
2.Identified that during real-time image generation, the values in the image were repeating, which affected the YOLO model’s detection of eye blinks. Currently exploring methods to convert 3.EEG signals into color images without value repetition in real-time.
4.Discussed with Mr. Sambhu ways to improve YOLO model accuracy. He highlighted that during training, some eye blink annotations were not accurate. In certain cases, a single eye blink with small peaks was incorrectly interpreted as two blinks by the model. The proposed solution is to select data from only one channel for eye blink detection during training for that during image creation time i will apply channel selection method and generate single channel image data for YOLO trainning.
5.Assisted Miss Dhanya in resolving a JSON configuration issue in the Filter Basic UI module during testing. The issue was successfully solved.
6.PL suggested integrating a common preprocessing module for all image generation functionalities — 80% completed, task in progress.
7.Visualizing live images from LSL stream data during YOLO testing in real-time — task in progress.
8.Explained the live event detection module and the simulator module to Miss Gourisree(front end working)-Back end code explanation pending.

Daily Report: October 23, 2024
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Integrated the preprocessing module into the image generation UI, focusing on the color image generation portion.
2.Identified that during real-time image generation, the values in the image were repeating, which affected the YOLO model’s detection of eye blinks. Currently exploring methods to convert 3.EEG signals into color images without value repetition in real-time.
4.Discussed with Mr. Sambhu ways to improve YOLO model accuracy. He highlighted that during training, some eye blink annotations were not accurate. In certain cases, a single eye blink with small peaks was incorrectly interpreted as two blinks by the model. The proposed solution is to select data from only one channel for eye blink detection during training for that during image creation time i will apply channel selection method and generate single channel image data for YOLO trainning.
5.Assisted Miss Dhanya in resolving a JSON configuration issue in the Filter Basic UI module during testing. The issue was successfully solved.
6.PL suggested integrating a common preprocessing module for all image generation functionalities — 80% completed, task in progress.
7.Visualizing live images from LSL stream data during YOLO testing in real-time — task in progress.
8.Explained the live event detection module and the simulator module to Miss Gourisree(front end working)-Back end code explanation pending.

Daily Report: October 24, 2024
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Addressed and resolved the configuration JSON errors reported by Miss Gourisree during the PyQt-based live real-time seizure EEG data run.
2.Resolved the issues related to anesthesia epoch generation and epoch data plotting.
3.Fixed a configuration JSON error reported by PL during deep learning raw data processing. After closing the UI, re-browsed the configuration JSON, resampling JSON, and channel selection JSON, which had not been set; this issue has been resolved.
4.Integration of a common preprocessing module for all image generation functionalities is in progress.
5.The task for single-channel eye blink-based YOLO model real-time detection is pending.
6.The task for visualizing live images from LSL stream data during YOLO testing in real-time is pending.

Daily Report: October 25, 2024
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Explained the MIME project structure to Miss Vaishnavi, covering the EEG data viewer, Filter UI, preprocessing packages, filter pipeline, preprocessing pipeline, Resampling UI, and its backend code. The explanation of the feature extraction section is pending.
2.Explained the live event detection module (pyqt based ) to Miss Gourisree(front end working)-progress bar implementation explained -Back end code explanation pending(thread part of the code)
3.The task for visualizing live images from LSL stream data during YOLO testing in real-time task in progress.

Daily Report: October 28, 2024
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Prepared a Word document detailing the ongoing tasks for the BMI team and forwarded it to PL for review.
2.Discussed the design and implementation of the epoch generation tool with FLA.
3.Guided Miss Lakshmi in creating the epoch generation tool within the EEG data viewer module and explained the menu class to her.
4.Participated in the discussion on the Icon-BCI conference schedule with FLA, SLA, and team members. SLA assigned me to update the expected participants' details in a excel sheet.
5.Successfully generated real-time images from stream data using MNE for image generation. Currently addressing a scaling issue where the EEG plot appears as a straight line, impacting model detection efficiency.

Daily Report: October 29, 2024
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Explained the general producer thread function and the consumer thread for EEG analysis to Miss. Gowrisree, with a focus on implementing a progress bar within the process. - Task completed.
2.FLA assigned me to cross-check the iConBCI site to ensure all delegates' photos and details are available, comparing them with the iConBCI conference Excel sheet. - Task completed (parallel task)
3.Miss. Anusree encountered an error during testing of SSVEP data with the trained model:
"File "C:\Users\IMRI\AppData\Local\Programs\Python\Python311\Lib\json\__init__.py", line 339, in loads raise TypeError(f'the JSON object must be str, bytes or bytearray, 'TypeError: the JSON object must be str, bytes or bytearray, not dict".
I resolved the issue, committed the updated code in SVN, and initiated retesting.
4.Assisted Miss Dhanya in resolving the key setting issue in the testUI JSON.
5.Modifying the conference paper based on the reviewer’s recommendations(• Eye blinks are essential non-verbal cues, conveying emotional states, attention, and intentions. EEG-based eye blink detection enables individuals with motor disorders or paralysis to interact with devices.
• Can EEG-based eye blink detection accurately distinguish between voluntary and involuntary eye blinks?
• How do eye blink patterns vary across different emotional states (e.g., stress, relaxation, attention)?
• Can machine learning algorithms effectively classify EEG signals into eye blink and non-blink events?)

Daily Report: October 30, 2024
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Integrated YOLO testing code into the live PyQt-based UI; task is in progress.
2.Corrected an error in MIME Version 3 (AI platform) - Task completed.
3.Guided Miss. Gowrisree on implementing the seizure progress bar in the PyQt UI and assisted Miss. Lakshmi with Python coding for auto-epoch creation from EEG data.

Daily Report: November 01, 2024
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Completed the task of correcting all deep learning algorithms to ensure they function based on the number of output classes configured through the GUI. Resolved the issue where some models required the output to be set to (number of output + 1) and ensured that the majority of the models now correctly operate with an output fixed at 32. Additionally, I also solved the plotting bin issue. These corrections were made in preparation for the demo scheduled for the upcoming Monday. Currently, we are testing these algorithms, which can dynamically detect unique classes from the labels themselves without requiring a predefined number of classes. These corrections and tests are in preparation for the demo scheduled for the upcoming Monday.
2.SLA provided the Water and Vellam event(inner speech) dataset with a 256 Hz sampling rate for upsampling to 4096Hz. I have completed the task of upsampling these data sets

Daily Report: November 04, 2024
----------------------------------------------------------------------------------------------------------------------------------------------------------
1.Explained the training module, general settings, and data upsampling/downsampling processes within the AI platform to SLA. Addressed SLA's questions regarding upsampling and downsampling techniques.
2.Attended Vigilance Awareness Week Program with theme: "Culture of Integrity for Nation's Prosperity".
3.Preparing the biodata for PI and Co-PI for the ICMR project proposal, following the specified ICNR project proposal biodata format-Task in progress

Daily Report: November 05, 2024
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Completed the biodata for FLA and SLA according to the ICMR proposal format.
2.Ongoing revision of the conference paper based on reviewers’ recommendations, focusing on the following points:

> The role of eye blinks as essential non-verbal cues, conveying emotional states, attention, and intentions; and how EEG-based eye blink detection enables interaction for individuals with
motor disorders or paralysis.
> Exploring the accuracy of EEG-based eye blink detection in distinguishing between voluntary and involuntary eye blinks.
> Analyzing variations in eye blink patterns across different emotional states (e.g., stress, relaxation, attention).
> Assessing the effectiveness of machine learning algorithms in classifying EEG signals into eye blink and non-blink events.
3.FLA reported an error in preprocessing module(High pass filter, Gaussian filter)-Pending Task
4.Alternate method for peak detection- Developing signal processing module for threshold detection with out using third-party package-currently used third-party module for that but not giving the expected output -Pending Task

Daily Report: November 06, 2024
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Prepared Dr. Gautham Arunachal Udupi’s resume from NIMHANS for the ICMR project proposal.
2.Completed revisions to the conference paper based on reviewers’ recommendations.
3.At FLA’s instruction, validated the epoch creation tool developed by Miss Lakshmi. Cross-checked and reported findings to FLA, and am guiding the tool’s development as per FLA’s guidance.
4.Addressed an error reported by FLA in the deep learning module while running EEGNet: “No positive samples in y_true, making the true positive value meaningless” - Pending Task.
5.Reviewed an error in the preprocessing module (High-pass filter, Gaussian filter) reported by FLA - Pending Task.
6.Developing an alternate method for peak detection. Currently creating a signal processing module for threshold detection without relying on third-party packages, as the current third-party
module is not providing the expected output - Pending Task.
7.Attended an LSTM presentation by Mr. Suraj.
8.Checked flight availability for the iConBCI project on December 19th (parallel task) and prepared word document

Daily Report: November 06, 2024
------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Prepared Dr.R Subasree resume from NIMHANS for the ICMR project proposal.
2.FLA reported errors such as " File "C:\Users\HSTG_MANICS\AppData\Local\Programs\Python\Python311\Lib\json\__init__.py", line 339, in loads raise TypeError(f'the JSON object must be str, bytes or bytearray,'TypeError: the JSON object must be str, bytes or bytearray, not dict" and File "C:\Users\HSTG_MANICS\AppData\Local\Programs\Python\Python311\Lib\site-packages\sklearn\utils\validation.py", line 778, in check_array dtype_orig = np.result_type(*dtypes_orig) File "<__array_function__ internals>", line 180, in result_type ValueError: at least one array or dtype is required".
3.Eye blink algorithm fine tunning started

Daily Report: November 11, 2024
-------------------------------------------------------------------------------------------------------------------------------------------------------
1.Resolved all reported errors in MIME version 3 software, including filter issues. High-pass filter (>45 Hz) and band-stop filter (45.5 Hz to 50 Hz) concatenation issues were successfully addressed.
2.Fine-tuning of the eye blink detection algorithm is in progress.
3.Resolved the data-saving issue in resampling, as reported by Miss.Dhanya.
4.Guided Miss Lakshmi in developing the epoch creation tool, including folder browsing and EDF data epoch segmentation, and explained the underlying logic.
5.Editing the conference paper according to reviewer comments (second revision, with 35% plagiarism reported) – task in progress

Daily Report: November 12, 2024
-------------------------------------------------------------------------------------------------------------------------------------------------------
1.Integration of emotive device with eye blink algorithm – task in progress.
2.Fine-tuning of eye blink algorithm completed; testing with simulator completed, real-time device testing pending.
3.Preparing eye blink detection system for GMCT demo with neurology doctors.

Daily Report: November 18, 2024
-------------------------------------------------------------------------------------------------------------------------------------------------------
1.Prepared the Minutes of Meeting (MoM) document for the discussion with GH on AI Platform Architecture, Organization, and Model Deployment (15-Nov-2024). The document was reviewed and forwarded to FLA.
2.Ongoing testing of the real-time eye blink detection algorithm using the Emotive device.
3.Completed real-time visualization of Emotive device data. Currently resolving an issue related to single-channel EEG data visualization.
4.Continuing the second revision of the conference paper based on reviewer comments, addressing a reported 35% plagiarism – task in progress.

Daily Report: November 19, 2024
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Following today’s morning discussion with FLA, he suggested some modifications to the eye blink detection code. The task involves visualizing all available channel data, but processing only the necessary channels (Fp1 or Fp2, which provide clear blink signals) in the background. This task is currently in progress.
2.Resolved installation issues and successfully assisted Miss. Dhanya with the installation of PyCharm and the Python interpreter on the new workstation (PC).
3.The second revision of the conference paper involved addressing the reviewer’s comments and correcting the reported 35% plagiarism. I modified the content as per the plagiarism report, ensuring the paper met the necessary academic standards. After completing these revisions, the updated paper was submitted on the conference management site for the third round of review.
4.Modified the resumes of Dr. Chithra (PI) and Dr. Praveen Panicker (Co-PI) in accordance with the ICMR project proposal format and shared them to FLA for review.
5.Prepared the resume of Byju Sir (PI) in the ICMR format and forwarded it to FLA for review.

Daily Report: November 20, 2024
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Modified ICMR-formatted resumes(identifying impact factor and H-index in research gate) of Dr. C.S. Mani (PI), Binu P.J. (CoPI), and Byju (CoPI), and forwarded updated PDFs to FLA for review. Combined Publication Details of PI and CoPI in table format-Task Completed.
2.Completed conference paper documentation, including e-copyright, PDF upload, and final camera-ready submission-Task completed.
3.Implemented channel selection(Coding) based on user-defined indices in MiMe software. Task completed.

Daily Report: November 21, 2024
---------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Resolved package installation issues for PySerial and PyLSL on the new laptop – Task completed.
2.Explored interfacing the Emotiv device with the new laptop – Task completed(parallel task).
3.Integrated a custom-made single-channel EEG device with the eye blink detection code and updated visualization functionality – Task in progress. Addressed a real-time eye blink detection 4.visualization issue reported by the FLA and working on smoother visualization for the EEG real-time visualizer-Task InProgress.
5.Assisted Miss. Gowresree in resolving an error encountered while plotting offline epochs of seizure data – Task completed (parallel task).
6.Upsampled volunteer data (256 Hz) provided by SLA, including "water" and "vellam" inner speech data – Task completed (parallel task).
7.SLA inquired why the upsampled file size for 4096 Hz data was "1.45 MB (15,28,987 bytes)" while the original 256 Hz data file size was "41.8 MB (4,38,73,587 bytes)" after upsampling using the NeuroKit Python package. I referred to the package documentation and provided an initial explanation. However, further exploration is needed to fully understand this discrepancy.
8.FLA reported an issue when traiining using ML SVM,IndexError: single positional indexer is out-of-bounds-Task InProgress (parallel task)

Daily Report: November 22, 2024
-----------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Resolved an error in the feature extraction functionality reported by FLA and Miss. Dhanya M.R., where the "Extract Feature Only" radio button caused issues with execution time. Attempted to commit the updated code to SVN, but encountered a commit error, which was reported to FLA.
2.Continued working on real-time EEG data visualization smoothing (task in progress).
3.Assisted Miss. Gowresree in resolving an error encountered while plotting offline epochs of seizure data – Task completed-Notch filter error also solved.
4.started the real-time threshold-based eye blink detection code integration with the the real-time eeg data from the hardware(CDAC-made) device task (task in progress).

Daily Report: November 25, 2024
---------------------------------------------------------------------------------------------------------------------------------------------------
1.Completed real-time EEG data visualization with proper smoothing (Task completed).
2.Integrated the real-time threshold-based eye blink detection code with real-time EEG data from the CDAC-made hardware device (Task completed).
3.Implemented a bandpass filter. Currently validating different combinations to optimize the eye blink detection algorithm (Task in progress).

Daily Report: November 26, 2024
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Eye Blink Detection: Addressed issues with stream oscillation and floating effects after applying a bandpass filter to the EEG stream. Explored solutions including padding techniques, adjusting filter order, and experimenting with different filter types such as Butterworth, FIR, and IIR. Further optimization in progress.

2.Fixed the issue reported by Miss. Dhanya regarding upsampling and downsampling in high-precision(it taking the data points like this "6.516600" ) in saved data. Task completed.

3.Prepared the November monthly report for the BCI team, which is intended for submission to MeitY. The Word document has been forwarded to FLA for their review before final submission.

Daily Report: November 27, 2024
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Finalized modifications to the November monthly report for the BCI team, which is intended for submission to MeitY. The updated Word document has been forwarded to FLA for review before final submission.
2.Eye Blink Detection: Resolved stream oscillation and floating effects. The implementation now allows dynamic adjustment of filter values while retaining existing filter functionalities. Integration of these offline filtering features into the online visualization module is currently in progress.
3.Incorporated FLA-suggested changes to the visualization module. Removed the box, vertical, and horizontal grid lines from the real-time plot. Task completed.
4.Explored key features for eye blink detection and trained an SVM classifier using data from Volunteer 5.

Daily Report: November 28, 2024
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Incorporated FLA's suggested modifications for eye blink visualization: added dotted x-axis and y-axis lines, trying to display time from 0 to 5 seconds on the plotting window, and implementing user alerts for single or double voluntary blinks within this time frame. Task in progress.

2.Resolved issues with the bandpass filter in real-time eye blink visualization, ensuring proper filtering and eliminating errors. Task completed.

3.Began offline analysis of eye blinks, finalized feature extraction parameters (Variance, Kurtosis, Skewness, Energy, Delta, Wavelet Energy, Root Mean Square, and Number of Zero-Crossings) using selected channels (Fp1, Fp2, A1, A2) and trained an SVM machine learning algorithm. Task in progress.

4.FLA reported an error in the MiMe version 3," File "D:\old PC\E\SVN Repository\HTG 047 D\BMI V 2 0\eBrainVoice_V3pt0\eBrainVoice\module\src\ai\ml\Classify.py", line 41, in Classifyclassify_json_extracted_from_settings_json = test_json["Settings"]["train_ai_technique"]["train_data_settings_json"]TypeError: string indices must be integers, not 'str'"

Daily Report: December 02, 2024
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Reported a critical issue during training where the training set was continuously repeating, leading to delays in the classification process. While debugging, identified an additional issue with the training results, where the UI was not closing properly. Debugging and resolution are ongoing.(70 % task completed)

Daily Report: December 03, 2024
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Resolved the issue of repeated testing and multiple result UI generation after closing the MiMe application. Also fixed the clearing of entry widgets on both sides, addressing the problem where some values did not change when switching from the training tab to the testing tab. Task completed.

2.Currently debugging and searching for a solution to the following error: "encountered in divide psd_norm = psd / psd.sum(axis=axis, keepdims=True)." Identified that the FFT value is 0 and working on an alternative solution for this issue. Task in progress.

Daily Report: January 07, 2025
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Evaluated Python codes (approximately 12) submitted for the BRAINATHON-25 EEG Attention Multiclassification Challenge, prepared a detailed Technical Analysis Report, and sent it to CDAC Delhi.
2.Provided an in-depth explanation of the web-based application to the team members.

Daily Report: January 08, 2025
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Updated Version 3 codes and integrated them into Version 4.
2.Addressed an issue reported by Vaishnavi where, after logging into the EEG data viewer, most functionalities of MiMe were not working. Resolved the issue and committed the code.

Daily Report: January 15, 2025
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Resolving the synchronization issue in the eye blink based assistive system algorithm

Daily Report: January 16, 2025
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Explored web application folder structure, reliability of the flask application, Readily Available web UI Pricing and design with team member Anusree and presented findings during a group discussion.
2.Identified the root cause of the synchronization issue in the eye-blink-based assistive system algorithm.

Daily Report: January 16, 2025
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Synchronization Issue Resolved: The synchronization issue was addressed by utilizing the distance parameter in the findpeak functionality. However, a phasing issue remains with detecting consecutive blinks. Currently exploring the calculation of sample distance in the findpeak function by referring to the scipy.signal.find_peaks documentation for better detection accuracy.
2.Team Meeting & EEG Processing Discussion: Attended the Anasthetiadepth Monitoring Project team meeting, where the discussion focused on real-time EEG visualization delay. Additionally, participated in the EEG Signal Processing in EEG Devices session, conducted by Mir. Ramesh Kumar.
3.Explored readily available Flask-based templates for purchasing with Miss. Anusree

Daily Report: January 27, 2025
--------------------------------------------------------------------------------------------------------------------------------
1.Integrated the simulator with the eye blink detection code and debugged the peak-finding functionality. Also tested with simulated data. The issue persists where a single peak, once detected, repeats. Resolving this issue in progress in eye blink detection algorithm.
2.Parallely resolved the .Mat file visualization issue in EEG data for anesthesia depth monitoring.

Daily Report: January 28, 2025
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Peak Repetitive Issue Resolution: Explored the documentation for scipy.signal.find_peaks to address the peak repetitive issue. Discovered that certain parameters were missing in the eye blink detection algorithm. Added these missing parameters, though the values for parameters like wlen, rel_height, and plateau_size are still being fine-tuned. Currently experimenting with different values for these parameters to optimize performance.
2.Parallely resolved the .csv file visualization issue in EEG data for anesthesia depth monitoring.
3.In classify module testing error reported solved the issue.
Daily Report: January 29, 2025
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Explored and gained a better understanding of how Speech-to-EEG and EEG-to-Speech architectures work.
2.Tested offline eye blink data collected from Natus and Emotiv using the current algorithm but observed differences in blink signatures. To address this, recording eye blink data from our
single-channel EEG device.
3.Writing code for EEG data recording, ensuring compatibility with both single-channel and multi-channel devices.(Task in progress)
4.Implementing this functionality in the EEG Data Viewer and testing it with the current eye blink algorithm to check if the repetitive blink issue persists. Task in progress.

Daily Report: March 20, 2025
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Integration of EMOTIVE device with real-time prediction module-Task in progress

Daily Report: March 25, 2025
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Integrated real-time plotting of Emotive EEG data within MIME software, enabling dynamic adjustments to window size, channel spacing, and sensitivity for enhanced visualization and analysis.
2.Attended LSTM class by Miss Aleena and Disease Modeling class by Mr. Prasidh

Daily Report: March 26, 2025
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Modified real-time plotting and visualization code – Initially, separate plot and analysis functions were set for Emotive and the simulator. After correcting the Emotive visualization part, a common real-time plotting and analysis function was implemented. (Task Completed)
2.Dynamic Filter Configuration – Users can now dynamically change the low-pass and high-pass filters. (Task Completed)
3.Dynamic Display Settings – Users can now switch dynamically between different visualization modes: real-time plotting, predictions, or both. (Task Completed)
4.Notch Filter Adjustment – Implementing dynamic notch filter adjustments. (Task In Progress)
5.Deep Learning Execution Issue Resolved – Previously, execution stopped after feature extraction when deep learning was selected. The issue was identified as JSON variables for the training UI not being set in the case of DL selection. This has been corrected. (Issue Resolved)

Daily Report: April 03, 2025
----------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Exported EEG Data: Successfully exported Mr. Tarun’s EEG EDF data (Sections 1 and 2) from the Emotiv device and copied it into the HTG share.
2.LSL Code Update: Integrated the updated encrypted data handling into the LSL code. Previously, there was a delay in processing encrypted data, which has now been resolved.
3.Hardware Handshaking: Implemented handshaking with the hardware, allowing the user to select a specific data type (encrypted or non-encrypted). This selection is sent to the designated port where the device is connected. Based on these instructions, the device dynamically adjusts its data type and sampling rate.
4.EEG Data Recording: The coding task for EEG data recording is currently in progress.
5.Explored the available Emotiv licenses and their features, specifically investigating whether the Standard License includes LSL interface support. Prepared and sent an inquiry email to the Emotiv support team for confirmation.

Daily Report: April 11, 2025
----------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.On the new PC after checking out our project code from SVN, an environment configuration issue occurred while trying to run it(SDK error) — resolved the issue and completed the setup.
2.Solved the issue with live SSVEP prediction — fixed the problem where EEG visualization was not working as expected during prediction.
3.Corrected the real-time EEG recording code.
4.Loaded 1 hour of EEG data into the EEG data viewer, displaying 10 seconds per window; also assisted with the integration of the Feature Extraction UI.

Daily Report: April 15-16, 2025
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Status of Meditation EEG Data Analysis ******************************************************************************************************************************************************************************
- resolved the issue with the delta frequency in the band frequency feature extraction, where the delta frequency was 0 across all channels.
- Implemented EEG feature extraction pipeline using Welch’s method.
- Loaded raw EEG data from EDF file and selected for 23 key EEG channels.
- Computed frequency domain features including Delta, Theta, Alpha, Beta, Gamma power and calculated band power ratios (Alpha/Beta, Delta/Theta, etc.).
- Combined all features by channel into a table and saved it as a CSV file for further analysis.
- currently the above feature extraction method is integrating with MiMe Feature extraction- Task in progress

Daily Report: April 17, 2025
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Integrated the channel-wise vertical feature extraction method into the MiMe software. This includes:
-Writing extracted features to CSV.
-Reading channels directly from the EDF file.
-Displayed all channels in the channel selection UI, allowing the user to select the desired channels. Integrated this functionality into the UI, committed all modified files to SVN, and
tested.
2.Generated line graphs based on the extracted features for visualization.
3.Explained the new feature extraction functionality to FLA and Ms. Lekshmi. **********************************************************************************************************
Meetings *************************************************************************************************************
-Attended a meeting with Subodh Sir regarding the installation of the standalone application for the CCET project.
-Participated in the iCon BCI Lecture Series (Part 9) on the topic: “The Utility of Whole Person Cognition for BCI.

Daily Report: April 21, 2025
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Explored and implemented GPU acceleration for CatBoostClassifier and XGBClassifier by enabling task_type='GPU' and tree_method='gpu_hist'.
2.Feature Extraction errors resolved and code committed in the SVN

Daily Report: April 22, 2025
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Completed batch-wise plotting of 1-hour seizure EEG data in MiMe software.-TASK COMPLETED

2.Modified GPU-based ML training code: logic added to automatically switch between CPU and GPU execution based on system capability. -TASK COMPLETED

3.Pending: Fix the Deep Learning (DL) error reported by the Project Lead. -PENDING TASK

Daily Report: May 21, 2025
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Feature extraction error correction and minor feature extraction ui updation, feature matrix creation diagram drawing and showing it as tooltip in ui, feature extraction type comparison
2.Overlap window implemented in MiMe, and testing.

Daily Report: June 03, 2025
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Configuration file was not updating with selected EEG channels – issue identified and fixed.
2.CNN-BiLSTM model's label map was not updating correctly – problem resolved.
3.Recorded EEG data for eye movement patterns (left, right, up, and down) for analysis.
4.Offline eye blink detection analysis code into the main pipeline.

Daily Report: July 07 to 11, 2025
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Corrected and integrated the following EEG features: Cross-Correlation, Decorrelation Time, Lyapunov Exponent, and Phase Locking Value,Made their parameters configurable through the UI.
2.Attended Axonent EEG device demonstration session.
3.Integrated BESS XAP (new Axonent EEG device) with the MiME software platform.
5.Completed code integration for seizure detection module, spinner functionality, and offline EEG data plotting, scalp map etc with the team.
6.Identified and resolved issues in the MiME Inner-Speech detection pipeline post integration.
7.In feature extraction output by including window size, overlap window, and patient ID in the saved files.

Daily Report: July 28 to Aug01, 2025
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Resolved SenSiz channel selection issue by handling cases without metadata files.-Task Completed
2.Fixed EDF file plotting and channel selection bugs in SenSiz.-Task Completed
3.Corrected EEG signal UI to support Multi-label module development.-Task Completed
4.Successfully implemented and completed multi-label creation module.-Task Completed and Testing pending
5.All modified codes have been committed to SVN and verified for proper functionality.

Daily Report: Aug 04, 2025
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Resolved multi-labeling issue in EEG Signal UI – Task Completed.

2.Fixed EDF file loading bug in SenSiz in collaboration with Vaishnavi – Task Completed.

3.Exploring and developing responsive UI for MiME software to adapt to changes in display settings (resolution, scaling) – Task Pending.

4.Debugging the feature extraction module to identify causes of slow performance – Task Pending.

Daily Report: Aug 05, 2025
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Fixed a bug in ExecutePipeline caused by the newly added multi-labeling functionality; identified the issue and resolved it.

2.Resolved the feature extraction slowness in Meditation EEG data.

Daily Report: Aug 18-22, 2025
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.When time and frequency domain features were combined, a reshape error occurred due to incompatible dimensions. This issue — caused by Delta Spectral Power — has been resolved, and the updated code has been committed to SVN.
2.Explored and implemented various channel data into deep learning algorithms without causing errors. The issue arose because channel selection was not initially integrated with the EEG Transformer model. This has been resolved by incorporating channel selection into the EEG Transformer pipeline.
3.Fixed: color, grayscale, and spectrogram image generation scripts. In progress: correction of epochs_To_grayscale_image.py and epochs_to_channelise_scalogram.py,scalp map generation.

Daily Report: Sept 15, 2025
-------------------------------------------------------------------------------------------------------------------------------------------------
1.Corrected the reported bug in MiME MindMe Version 2 software-Testing pending.
2.Preparing the MiME MindMe software for the upcoming 19th demo.

Daily Report: oct 6-12, 2025
-------------------------------------------------------------------------------------------------------------------------------------------------
1.Official tour to Purple Fest 2025.
Daily Report: oct 13-14, 2025
-------------------------------------------------------------------------------------------------------------------------------------------------
1.Attended discussions on web-based model creation, execution module development, and backend integration.
2.Processed 4096 Hz simulator data in MiMe visualization; initially appeared jumpy due to high resolution—issue resolved.
3.Updates to MiMe UI as suggested by GH are in progress.
4.Explained inner speech decoding protocol and EEG data naming conventions to juniors.
5.Sent follow-up emails to all contacts made during Purple Fest 2025 and arranged a meeting with NEMA AI.

Daily Report: Nov 14, 2025
-------------------------------------------------------------------------------------------------------------------------------------------------
1.In the deep learning–integrated image-based training and testing module, the graph-saving functionality has also been integrated

Daily Report: Nov 17, 2025
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Prepared the Memorandum of Understanding (MoU) for the project titled “The MCI Connectome: Mapping Brain Connectivity to Predict Dementia Conversion Risk in Subjects with Mild Cognitive Impairment and Subjective Cognitive Decline – Blood Biomarker-Based Validation.”
The document has been forwarded for review. (Task completed.)

2.Integrated the web-based model execution module as part of the ongoing system development workflow.
(Task in progress.)

Daily Report: Nov 24, 2025
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
When running our Tkinter-based Python UI application on laptop, the program crashes during real-time EEG streaming with the following error:
RuntimeError: main thread is not in main loop
Tcl_AsyncDelete: async handler deleted by the wrong thread

Currently resolving this issue

Daily Report: Nov 26, 2025
---------------------------------------------------------------------------------------------------------------------------------------------------------------
1.I have resolved the reported issue related to the Tkinter-based Python UI crashing during real-time EEG streaming, which previously showed:
-RuntimeError: main thread is not in main loo
-Tcl_AsyncDelete: async handler deleted by the wrong thread
After debugging, I corrected the thread-handling logic and tested other related settings. A few additional issues were identified during testing, and those have also been fixed. The application is now running smoothly without crashes.

Daily Report: December 18, 2025
---------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Resolved an issue in real-time EEG stream saving where certain EEG channels were missing during data storage.
2.Prepared the MiME software for the inner speech demo
3.Modified the Proposal for Signing MoU with SCTIMST document as instructed and forwarded it for verification.

Daily Report: December 19, 2025
---------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Integrated the attention detection function into the updated codebase; identified issues after SVN update and resolved them.
2.Resolved Python interpreter configuration issues in PyCharm on a team member’s PC.(Modified requirement text file)
3.Cross-checked and verified the MiME patent document.
4.Prepared a step-by-step operating guide for MiME software (real-time and offline modes) and uploaded it to the HTG shared repository. Additionally, uploaded the working software code, AI models, configuration files, and test datasets to the HTG share for future demo purposes.

Daily Report: 02-01-2026
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

1.As part of the Anesthesia Depth Monitoring project, performed EEG hardware–software setup and interface configuration at SCTIMST, Thiruvananthapuram for EEG data collection. Software interface issues were identified and resolved, and a junior doctor was familiarized with the EEG system and software operation.

2.Resolved a TypeError crash in execute_pplfunc caused by uninitialized resampling settings when the Resampling UI was opened and closed without selection.

3.Identified a Tkinter callback error during pipeline execution caused by non-unique EEG channel selection in the feature extraction stage (pick_channels ValueError). Analysis and correction of the channel-selection logic are currently in progress.

Daily Report: 15-01-2026
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Recorded inner-speech EEG signals using the Emotiv device for the events Water, Vellam, Food, Aharam, Left, Right, Edathottu, and Valathottu. The collected data was provided for training and tested with the real-time MiME MindME software.
2.Preparing the MiME MindME software for demonstration at SRM, with a focus on real-time EEG-based inner speech decoding.
3.Prepared a Word document for the MiME software brochure and shared it with Anusree for review.

Daily Report: 27-01-2026
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Completed image-based deep learning model training and testing; committed the updated code to SVN.

2.Re-training functionality integrated using the re-training code shared by Hemal and added into the AI module.

3.Akhila C++ code integration pending

Daily Report: 09-02-2026
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Prepared a PowerPoint presentation outlining BCI project activities under the Health Technology Group for presentation to a visiting delegate from Brazil( Dr. Jorge Edison Ribeiro (Aracaria Foundation)). The PPT has been shared with Binu Sir for review.
2.Algorithm development for visual event marking and annotation of seizure events in EEG data; work is currently in progress.
3.Prepared the Declaration Certificate for C-DAC for the proposal titled “Development of Indigenous Hypotension Prediction Index (HPI) using an AI-based algorithm for perioperative hemodynamic assessment and optimal management in non-cardiac surgical patients.” Budget preparation is in progress.
4.Presented BCI project activities under the Health Technology Group to Dr. Jorge Edison Ribeiro (Aracaria Foundation).

Daily Report: 10-02-2026
--------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Visited Sree Chitra Tirunal Institute for Medical Sciences and Technology (SCTIMST), Thiruvananthapuram, to resolve issues with the EEG device deployed in the ICU for anesthesia-related EEG data recording. Identified an issue with the SPU monitor and cleared client and server temporary files to address storage constraints on the C drive. The attending doctor tested the system to verify whether the issue persisted.

2.Prepared the budget allocation for C-DAC Thiruvananthapuram for the proposal titled “Development of Indigenous Hypotension Prediction Index (HPI) using an AI-based algorithm for perioperative hemodynamic assessment and optimal management in non-cardiac surgical patients.” Budget preparation completed. Prepared ED approval form also, completed declaration form , modified proposal also

Daily Report: 11-02-2026
--------------------------------------------------------------------------------------------------------------------------------------------------------------------
Leave
Daily Report: 12-02-2026
--------------------------------------------------------------------------------------------------------------------------------------------------------------------
Design and implementation of an algorithm for automated visual event detection and seizure annotation in EEG datasets; development in progress.

Daily Report: 12-02-2026
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
• Continued implementation of the multiple labeling algorithm (development in progress).
• Prepared the project registration document for The MCI Connectome: Mapping Brain Connectivity to Predict Dementia Conversion Risk in Subjects with Mild Cognitive Impairment and Subjective Cognitive Decline – Blood Biomarker-Based Validation.
• Prepared the brief project write-up (Word document) for MBM.

Daily Report: 17-02-2026
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Prepared project registration document and project write-up for the PNB project.
2.Attended an online meeting with the AIIMS Raebareli team.
3.Resolved the issue occurring while switching from Graph radio to Text radio, where text-based multi-label components were not clearing properly. Created a side panel to list all available EEG files selected during data loading; clicking on a file now plots it. Fixed EDF file plotting issues. Currently working on the annotation function and automatic event display from event.json (development in progress).
4.Correction in the MBM & PNB registration document

Daily Report: 24-02-2026
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Integrated EEG visualization–based event marking for the seizure dataset with epoch marking, feature extraction, and the AI pipeline. The major portion of the algorithm development has been completed. Testing is currently in progress. The feature for adding new events within the event marking module is pending.

2.Conducted real-time testing of the MiME (MindMe – Inner Speech Decoding) application using the Axxonent EEG device at SCTIMST, Thiruvananthapuram.

Daily Report: 25-02-2026
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Conducted real-time testing of the MiME (MindMe – Inner Speech Decoding) application using the Axxonent EEG device at SCTIMST, Thiruvananthapuram.

Daily Report: 27-02-2026
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Outdoor Duty at SCTIMST, Thiruvananthapuram
1.Implemented the method suggested by the Axxonet team to enable EEG LSL data access on another laptop using the Sentinel key provided by them. Configured the LSL streaming application and SPU Monitor for data acquisition.

2.Tested the MiME (MindMe – Inner Speech Decoding) application with the Axxonet EEG device. Simultaneously collected inner speech data for further analysis.

Daily Report: 07-03-2026
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Prepared a project proposal titled "EEG-Based Neural Representation of Imagined Speech Syllables for Assisted Communication in the Indian Population" for submission to the Indian Council of Medical Research (ICMR).

2.Exploring techniques for inner speech decoding using EEG signals.

3.Parallell worked on drafting a journal paper on inner speech decoding.

Daily Report: 11-03-2026
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Debugged and resolved a shape mismatch issue in the offline EEG moving window visualization by aligning the time vector and processed EEG data dimensions before plotting.
2.fixed the channel labeling to display the actual selected channel names from the UI instead of numeric indices
3.Completed PyCharm setup on the new laptop, and the MiMe interpreter configuration is currently in progress.

Daily Report: 12-03-2026
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Inner Speech Decoding: EEG visualization with event display is working properly on the new laptop. No threading issues were observed during testing. All Python interpreter issues on the new laptop have been resolved.

2.Feature Development: Implemented code to allow displaying signals only, events only, or both together. Currently exploring methods to achieve this without storing data in memory.

Daily Report: 13-03-2026
----------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Code Integration and Debugging: Resolved issues related to EEG plot display when showing signals alone or together with events. The code has been integrated for both inner speech real-time detection and seizure detection. Several code-level issues were addressed, including Filter State Shape Mismatch Error and thread-related issues. Testing has been completed. Code commit to SVN is pending.

2.Data Processing and Model Training: Combined volunteer and patient datasets and resampled the data to 1000 Hz. The Random Forest classifier training for inner speech decoding has been initiated on the new laptop for real-time inner speech decoding.

3.Research Exploration: Exploring new techniques in inner speech detection to improve real-time decoding accuracy.

Daily Report: 16-03-2026
----------------------------------------------------------------------------------------------------------------------------------------------------------------
Integrated the eye-blink threshold–based assistive communication algorithm into the new framework. Some issues were observed in the processing logic during testing, and debugging is currently in progress.

Daily Report: 01-04-2026
----------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Completed the inner speech decoding analysis and generated results using datasets tested under various conditions. Prepared result sections for Offline Inner Speech Decoding: AI Model Test Analysis in Offline Data – Results and Inner Speech Decoding: AI Model Test Analysis – Simulator-Based Results.
2.Added the latest technical details, software screenshots, and analysis results to the 4th Closure PRSG PPT.

Daily Report: 05-06-2026
----------------------------------------------------------------------------------------------------------------------------------------------------------------
1.Added results table printing to all 9 DL models (ANN, LSTM, RNN, GRU, EEGNet, BiLSTM, BiGRU, CNN-LSTM, CNN-BiLSTM) — metrics include Accuracy, Precision, Recall, F1, AUC-ROC, Loss, and # Parameters. Each model now stores results in self.results for the journal comparison table.
2.Created a structured Excel sheet to record inner speech DL results (ANN, LSTM, RNN, GRU, EEGNet, BiLSTM, BiGRU, CNN-LSTM, CNN-BiLSTM) with columns for all key metrics — Accuracy, F1, AUC-ROC, Loss, and Parameters — organised across 4 sheets for journal reporting.
3.Deep learning model code integration into the web application is currently in progress, including updates to support all nine architectures