Researchers propose sheaf hypergraph network for GBM survival prediction
A new study on arXiv introduces a multi‑modal framework for predicting glioblastoma multiforme (GBM) survival that aims to be both accurate and interpretable. The approach combines three components: a sheaf hypergraph neural network that models higher‑order relationships among MRI tissue patches using directional, asymmetric message passing; a concept bottleneck layer that forces learned…
Key points
- Sheaf hypergraph neural network captures higher‑order tissue patch relationships via directional asymmetric message passing.
- Concept bottleneck layer compresses representations into clinically grounded concepts for ante‑hoc interpretability.
- Model achieves 0.643 concordance index on 593 UPenn‑GBM patients with 0.015 standard deviation across folds.
The authors evaluated the framework on 593 patients from the UPenn‑GBM dataset using 5‑fold cross‑validation. Their model achieved a concordance index of 0.643 with a low fold‑level variance (standard deviation = 0.015), outperforming compared baselines. The paper claims to be the first to integrate sheaf hypergraph convolution, concept bottleneck supervision, and EST regularization for interpretable brain MRI survival prediction.
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Research
All →- Semantic Routing Calibration mitigates LLM over-refusal · 1 src
- AIBuildAI-2.5 ranks first on MLE-Bench with 73.3% medal rate · 1 src
- Researchers fine-tune 406M model for meeting summaries with retrieved text spans · 1 src
- Researchers test how language models handle numerical formats in word problems · 1 src
- Author pretrains language model End-to-End in Rust for $164 · 1 src
Comments
via GitHub Discussions