Summary of Tacco: Task-guided Co-clustering Of Clinical Concepts and Patient Visits For Disease Subtyping Based on Ehr Data, by Ziyang Zhang et al.
TACCO: Task-guided Co-clustering of Clinical Concepts and Patient Visits for Disease Subtyping based on EHR Data
by Ziyang Zhang, Hejie Cui, Ran Xu, Yuzhang Xie, Joyce C. Ho, Carl Yang
First submitted to arxiv on: 14 Jun 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
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 The proposed framework, TACCO, addresses the limitations of existing disease risk prediction methods by jointly discovering clusters of clinical concepts and patient visits in Electronic Health Records (EHR) data. By developing a self-supervised co-clustering framework guided by specific diseases’ risk prediction tasks, TACCO demonstrates significant performance improvements over traditional machine learning baselines. The model’s hypergraph architecture is enhanced with textual embeddings and contrastive objectives to enforce alignment between clusters of clinical concepts and patient visits. Comprehensive experiments on the MIMIC-III and CRADLE datasets show an average 31.25% improvement in phenotype classification and cardiovascular risk prediction tasks, with a 5.26% gain over the vanilla hypergraph model. The paper’s results are further validated through in-depth model analysis, clustering results, and clinical case studies. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary TACCO is a new way to predict diseases using electronic health records (EHRs). Existing methods don’t account for different subtypes of complex diseases. TACCO helps by finding groups of medical concepts and patient visits that are related to each other. This framework can be used to improve disease risk prediction, especially for cardiovascular diseases. The results show that TACCO is more accurate than traditional methods, which is important for making good decisions about patients’ health. |
Keywords
* Artificial intelligence * Alignment * Classification * Clustering * Machine learning * Self supervised
