GSOC 2026 Project #21 : I-EEG: health application for semi-automated iEEG analysis and viewing in a clinical and research environment

GSoC
1

Mentors: Maya Aderka <maya.aderka@mail.mcgill.ca>, Suresh Krishna <suresh.krishna@mcgill.ca>, Elie Bou Assi <elie.bou.assi.chum@ssss.gouv.qc.ca>

Skill level: Intermediate – Advanced

Required Skills: Fluency in at least one of Python or Matlab (Python preferred, and ideally with reasonable ability in MATLAB). Experience with (bio)signal processing and with front-end development preferred.

Time commitment: Full time (350 hours)

About: There is an acute need for (open-source) software to handle human intracranial neural recordings, usually from patients who are undergoing diagnostic intracranial EEG (iEEG) recordings for epilepsy treatment. Such software would provide an integrated viewer, implement major existing (semi-)automated algorithms for seizure-onset zone definition, seizure prediction and surgical outcome prognostication. This is a fairly new project, that the GSoC contributor will build on, with help and mentorship from us.

Aims: This year, the project will aim to implement additional key published iEEG algorithms, improve the front-end for visualization, create a pipeline that allows the incorporation of expert human input to fine-tune the automated analyses, and to build a framework for testing/comparison of different algorithms and human-in-the-loop pipelines to each other.

Website: GitHub - m2b3/ieeg · GitHub

Tech keywords: Health ML/AI, Epilepsy, Neuroscience, Biosignals, Neuroinformatics

2

Hello Maya, Suresh, and Elie,

I am Sai Krishna Vamshi, a first-year MTech student in AI at IISc Bangalore. My advisor is affiliated with the Centre for Neuroscience, and my master’s dissertation starting in August 2026 is planned around EEG-based seizure prediction.

I found the I-EEG project very interesting because it closely matches the area I want to work in seizure analysis, signal processing, neuroscience, and practical open-source tools for healthcare and research. I was also excited to see that this is a relatively new project since the opportunity to contribute at an early stage and help build something meaningful feels very motivating to me.

My current background is stronger in AI/Python, and I am actively learning more about EEG datasets, seizure-analysis workflows, and related literature. Since my dissertation will also be in a related direction, I feel this project is very well aligned with my academic interests.

I am reaching out before the application opens because I wanted to prepare properly and understand how I could focus my efforts on the right direction. I had a few questions regarding the project:

  1. For someone preparing seriously for this project, what would you suggest focusing on first?
  2. When the description mentions implementing additional key published iEEG algorithms, does that mean there are already some algorithms implemented and the contributor would extend that work?
  3. When it mentions improving the front-end for visualization, is there already an existing visualization interface or prototype that the contributor would build on?
  4. If there are already some implementations or prototypes for these parts, would it be possible to share them or briefly describe the current state? That would help interested students prepare in a more informed way before applying.

Thank you for your time. I am genuinely interested in this project and would greatly appreciate any guidance on how to prepare well.

Best regards,
Sai Krishna Vamshi

1 Like
3

Thank you for your interest, you are welcome.

You can join the iEEG community at alphatest.scicommons.org.

The key focus is on implementing and validating published algorithms for seizure onset zone identification and seizure prediction.

There is a prototype GUI that is actively being built and that will be ready by the time the GSoC coding period starts.

1 Like
4

Hello Suresh,

Thank you for the information. I will join the iEEG community on alphatest.scicommons.org and continue following the discussions there.

Best regards,
Sai Krishna Vamshi

1 Like
5

Hello Suresh, Maya, and Elie,

My name is Chenfan Liao, an undergraduate student majoring in Intelligent Medical Engineering.

I am extremely interested in this project. My current research directly focuses on the nonlinear dynamic analysis of epilepsy sEEG data (using Koopman operator theory), so I am very familiar with seizure-onset zone definitions and clinical EEG workflows. Additionally, I have extensive full-stack web development experience, making me well-prepared to improve and build upon the prototype GUI you mentioned.

I have just requested access to the iEEG community on alphatest.scicommons.org as advised. I will also follow up via email shortly to discuss some specific technical details for my GSoC proposal. Looking forward to contributing!

Best regards,
Chenfan Liao

6

@CHENFAN_LIAO - please use neurostars and/or alphatest for communication, rather than email. and thank you for your interest. your web development skills and importantly, your experience with seeg data seems highly relevant, so we welcome your participation and contributions and look forward to your proposal.

As always, the proposal should be about what you want to do, why, how, when and why you should be the person to do it. please take a look at the incf gsoc proposal template as well at: Recommendations for GSoC contributors | INCF

1 Like
7

Thank you for the warm welcome, Suresh! I’m glad to hear that my background in sEEG and full-stack development aligns well with the I-EEG project. I will strictly use Neurostars and the Alphatest community for further communications as suggested.

I have already started reviewing the INCF proposal template and will share a draft of my project plan on Alphatest soon for your feedback. Looking forward to contributing to the community!

8

@CHENFAN_LIAO - you can share it here via DM

1 Like
9

Got it! Thank you, Suresh. I will send you the Google Doc link via DM as soon as my initial draft is ready.

10

Hi Maya and Drs Suresh & Elie,

I’m Omar, a senior computer engineering & software systems design student at Ain Shams University in Cairo.

I have solid experience building end-to-end EEG-based BCI systems. I’ve developed a mobile app for real-time, “in the wild” EEG emotion detection, with focus on real-world deployability and a user-friendly UI for non-expert users.

I’ve also run a comprehensive suite of ML/DL experiments for different EEG classification tasks, so I’m highly aware of common pitfalls like data leakage, class label imbalance, and overfitting.

I have a strong software engineering foundation, so I’m confident I can contribute effectively to building a rigorous testing framework for the different algorithms, help build an intuitive and visually pleasing UI, and engineer a pipeline that seamlessly incorporates different algorithms and human input. All while making sure the system is modular, maintainable, and extensible.

To prepare, I’ve been studying some of the major seizure prediction algorithms and Dr. Elie’s work on connectivity features. I’ve also downloaded the Omni-iEEG dataset and I’m currently writing a local Python pipeline to implement these algorithms and test their accuracy against the expert annotations.

I have a question about the prototype GUI currently in development: What is the division of labor between the frontend and backend for data ingestion and algorithm integration? Specifically, is the GUI designed to ingest the raw clinical files (BIDS/.edf) directly into memory as a “thick client”, or will a decoupled Python backend handle the heavy parsing and serve the data chunks and algorithm outputs to the frontend via an API/WebSocket?

I know many clinical tools rely on thick client architectures, while a decoupled backend can sometimes offer more UI responsiveness. I’m comfortable working with either approach, but knowing your current architecture will help me properly scope the technical deliverables in my proposal.

I’m really excited to contribute to this project as it perfectly aligns with my dedication to neurotechnology and my recent BCI work.

Thanks.
Omar

11

@Omar thank you for your interest, and the initial work that you have done. I recommend that you also go through the messages above, if you have not already.

Regarding your question, currently, it is a thick client, but these are early days, and we may well combine the two approaches. In any case, you can assume one of the two when writing your proposal - given that you are writing the proposal blind with about 10 days left for the deadline, the deliverables written in your proposal are not binding, and are meant to be illustrative and a demonstration of vision, skills and scope. They will likely change during the course of the project.

Good luck !

1 Like
12

Sounds good! Thanks so much for the info.

I’ll write up my proposal draft and DM it to you in the next couple of days.

Thanks!

13

Hello Maya, Suresh, and Elie,

I hope this message finds you well. I apologize for joining the discussion a bit late, but after reviewing the project specifications, I’m confident that this project aligns perfectly with our shared objectives. As a neuroscience undergraduate with extensive engineering coursework, I am deeply passionate about neurotechnology and driven by clear architectural and creative visions.

My current research in a Biomedical Electronics and Bioinformatics lab, where I analyze fMRI and EEG data for generalized epilepsy (PTZ-induced) and therapeutic focused ultrasound (FUS) interventions, has enabled me to simultaneously write a computational neuroscience paper focused on functional connectivity and ML frameworks. My work involves approaching seizure onset at a macroscopic network level, shifting from the current paradigm that focuses on localized activities, and using multi-layer metrics to pinpoint seizure inhibition timing with maximum precision. I am eager to build your application not just to spec, but to seamlessly handle real-world, high-noise pathological data for extension into various neurological disorders.

Regarding your technical requirements:

  • Python, MATLAB & Biosignal Processing: Python is the backbone of my neuroimaging and EEG data pipelines. I use MATLAB for MRI data processing and visualization, such as generating dynamic connectivity matrices/heatmaps. Analyzing pathological brain states and training lightweight ML models are core parts of my lab work and BCI projects, too. I am also enhancing my ML and DL capabilities with the guidance of a renowned ML faculty member in my country. Additionally, I am highly proficient in using Excel for data analysis, such as functional connectivity dynamics.

  • Front-End Development: I enjoy building native applications. Designing a sleek, intuitive, and highly functional UI that serves clinical workflows is second nature to me. Most of my applications are neurotech-focused, featuring neuroadaptive designs and BCI-ready integrations.

I have reviewed the thread and want to ensure I stay on track and propose features within your scope. I would like to align my proposal and my preliminary contributions exactly with your current roadmap.

Looking forward to collaborating.

Best regards,

Silvie

1 Like
14

Welcome. Please go through all the messages above.

1 Like
15

Hello Maya, Suresh, and Elie,

I’m Krishnanshu Mittal from Delhi, India.

I’m a software engineer at Upside Down Labs, where I work with open-source bio-potential signal acquisition hardware and firmware. My work includes real-time EEG and EMG pipelines with custom IIR filters, signal processing, FFT-based spectral analysis, brainwave band extraction (delta, theta, alpha, beta, gamma) and focus detection from EEG signals, and ECG-based physiological monitoring including R-peak detection and breathing patterns.

I’m applying for GSoC 2026 Project #21 - iEEG, as it directly connects to my hands-on experience in biosignal processing, algorithm implementation, and data visualization for physiological signals. I’m particularly interested in contributing to the seizure onset zone detection algorithms and the human-in-the-loop pipeline for expert-guided analysis. I’ve also introduced myself on the BreathState thread (Project #20) and am actively exploring both projects.

I have gone through the thread and I’m currently exploring the iEEG repository and the Omni-iEEG dataset while studying published seizure onset zone detection algorithms to prepare my proposal.

GitHub
Portfolio

Regards,
Krishnanshu Mittal

1 Like
16

@clumsy, you are welcome. looking forward to the proposal.

17

Hi MENTORS,

I’ve submitted my GSoC proposal for the iEEG project and wanted to briefly express my strong interest. I’ve been consistently working with neurotech systems, especially EEG/EOG-based applications, including research papers and conference work in this domain. I also have hands-on experience with biosignal processing through projects involving real-world data and hardware (including work inspired by Upside Down Labs systems).

For this project, I have deeply studied the repository, discussions, and existing challenges. I am particularly interested in contributing to improving algorithm pipelines and building a robust human-in-the-loop framework for iEEG analysis. My current research direction—converting EEG signals into meaningful representations (text-level interpretation)—aligns closely with the long-term vision of intelligent neural data systems.

I would be very grateful if you could take a look at my proposal and share any feedback or direction for improvement.

Thank you for your time and consideration.

Best regards,

Lavina Korani.

GSOC PROPOSAL DOCS

1 Like
18

you only need to give visibility to my email address. as for your proposal and comments, i recommend that you add more detail regarding the experience you mention, for example signal processing etc. At this point, i have no additional feedback. All the best !

19

Dear Sir,

Thank you very much for your guidance. I have incorporated the suggested changes into my proposal, expanded my experience section in detail, and added links to my research paper, review papers, and GitHub projects.

Before I begin contributing, I would really be grateful for your advice on the preferred starting point for contributions. Could you please suggest which areas or tasks you would recommend I begin with?

Looking forward to your guidance.

Thank you again for your time and support.

Best regards,
Korani Lavina

20

We will start engaging with whoever is still around after the GSoC results are announced (selected intern(s) and volunteers).

1 Like