Skip to content

Add MV-AFA seizure detection algorithm#89

Open
yuhangYH wants to merge 6 commits into
esl-epfl:mainfrom
yuhangYH:add-mv-afa
Open

Add MV-AFA seizure detection algorithm#89
yuhangYH wants to merge 6 commits into
esl-epfl:mainfrom
yuhangYH:add-mv-afa

Conversation

@yuhangYH

@yuhangYH yuhangYH commented Jun 12, 2026

Copy link
Copy Markdown

Algorithm

MV-AFA (Multi-View Adaptive Fusion Attention) — topology-aware EEG seizure detection fusing four views of each 2-second window (temporal Transformer, frequency 2-D CNN, statistical features, TDA persistent homology) via a gated mixture.

v1.3.0 — multi-dataset (expanded TUSZ)

Trained across CHB-MIT + Siena + TUH-Sz (TUSZ, 1500 recordings) with subject-level splits, EEG augmentation and per-dataset balancing. SzCORE scores event-F1 across 5 datasets, so a model that has seen multiple corpora beats a single-dataset specialist; leave-one-dataset-out experiments confirmed zero-shot transfer to an unseen corpus is near chance, so the model is trained on each target corpus. Referential 10-20 inputs are auto-remontaged to 18 bipolar pairs.

Expanding TUSZ training data (400→1500 recordings) improved held-out window AUROC across all datasets (overall 0.732→0.745, TUSZ 0.658→0.675).

Docker image

docker.io/mellow99/mv-afa-szcore:v1.3.0

Training datasets

CHB-MIT, Siena, TUH-Sz (TUSZ)

Repository

https://github.com/yuhangYH/mv-afa-szcore

Author

Yuhang Guo — Khalifa University

@yuhangYH

Copy link
Copy Markdown
Author

Hi @jon-dan, thanks for maintaining SzCORE! This PR adds MV-AFA — the Docker image (docker.io/mellow99/mv-afa-szcore:v1.3.0) is published and tested end-to-end on EDF input, producing valid BIDS TSV output. Whenever you have a chance to run it on the benchmark, that'd be great — happy to address anything needed from my side. Thanks!

@jon-dan

jon-dan commented Jun 19, 2026

Copy link
Copy Markdown
Member

Hello, thank you for your contribution!

I cannot pull the image (see CI logs)

Can you check:

  1. Image is publicly available
  2. The image is built for linux/amd64

@yuhangYH

Copy link
Copy Markdown
Author

Thanks @jon-dan, and sorry about that! Both issues are now fixed:

  1. Public — the Docker Hub repo mellow99/mv-afa-szcore is now public (it had defaulted to private).
  2. Platform — the image was accidentally built for arm64 (I built on Apple Silicon). I've rebuilt and pushed mellow99/mv-afa-szcore:v1.3.0 for linux/amd64.

You can verify:

docker manifest inspect mellow99/mv-afa-szcore:v1.3.0   # -> linux/amd64
docker pull --platform linux/amd64 mellow99/mv-afa-szcore:v1.3.0

Should be good to re-run on the benchmark now — thank you for your patience!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants