Summary of A Self-supervised Model For Multi-modal Stroke Risk Prediction, by Camille Delgrange et al.
A Self-Supervised Model for Multi-modal Stroke Risk Prediction
by Camille Delgrange, Olga Demler, Samia Mora, Bjoern Menze, Ezequiel de la Rosa, Neda Davoudi
First submitted to arxiv on: 14 Nov 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This study presents a novel self-supervised multimodal framework that combines 3D brain imaging, clinical data, and image-derived features to improve stroke risk prediction prior to onset. The framework is based on contrastive learning and leverages large unannotated clinical datasets to capture complementary information across modalities. The model outperforms state-of-the-art methods in ROC-AUC (2.6%) and balanced accuracy (3.3% and 7.6%). Interpretable tools demonstrate better integration of tabular and image data, providing richer embeddings. The study showcases a robust self-supervised multimodal framework for stroke risk prediction, with potential applications in clinical predictive modeling. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper introduces a new way to predict the risk of having a stroke by combining different types of data, like brain scans and medical records. The approach uses a special type of learning called contrastive learning that helps machines understand how different kinds of data relate to each other. This method is trained on a large dataset from the UK Biobank and performs better than other methods for predicting stroke risk. |
Keywords
» Artificial intelligence » Auc » Self supervised
