Prototype Learning to Create Refined Interpretable Digital Phenotypes from ECGs

2025-08-02

AI Summary

This study investigates whether prototype learning can automatically discover clinically consistent and interpretable digital phenotypes from electrocardiograms (ECGs). Method: A prototype neural network was trained for multi-label ECG classification on the PTB-XL dataset; the learned prototypes were then transferred to the MIMIC-IV cohort and evaluated for association with discharge diagnoses via PheCode mapping. Results: Prototypes significantly outperformed conventional classification outputs and NLP-extracted concepts, demonstrating strong, disease-specific clinical associations—particularly for atrial fibrillation, heart failure, sepsis, and renal disease (AUC 0.89–0.91). Crucially, intra-prototype distance served as a quantitative indicator for phenotypic refinement, enabling disentangled identification of latent clinical features directly from physiological signals. This work provides the first evidence that purely classification-driven, unsupervised semantic-guided prototypes can spontaneously capture cross-dataset, multi-organ-system clinical structures with intrinsic interpretability.

Technology Category

• Machine Learning: Bio-inspired Learning
• Natural Language Processing: Language Grounding & Multi-modal NLP
• Computer Vision: Multi-modal Vision

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Prototype-based neural networks offer interpretable predictions by comparing inputs to learned, representative signal patterns anchored in training data. While such models have shown promise in the classification of physiological data, it remains unclear whether their prototypes capture an underlying structure that aligns with broader clinical phenotypes. We use a prototype-based deep learning model trained for multi-label ECG classification using the PTB-XL dataset. Then without modification we performed inference on the MIMIC-IV clinical database. We assess whether individual prototypes, trained solely for classification, are associated with hospital discharge diagnoses in the form of phecodes in this external population. Individual prototypes demonstrate significantly stronger and more specific associations with clinical outcomes compared to the classifier's class predictions, NLP-extracted concepts, or broader prototype classes across all phecode categories. Prototype classes with mixed significance patterns exhibit significantly greater intra-class distances (p $<$ 0.0001), indicating the model learned to differentiate clinically meaningful variations within diagnostic categories. The prototypes achieve strong predictive performance across diverse conditions, with AUCs ranging from 0.89 for atrial fibrillation to 0.91 for heart failure, while also showing substantial signal for non-cardiac conditions such as sepsis and renal disease. These findings suggest that prototype-based models can support interpretable digital phenotyping from physiologic time-series data, providing transferable intermediate phenotypes that capture clinically meaningful physiologic signatures beyond their original training objectives.

Problem

• Assess if ECG prototypes align with clinical phenotypes
• Evaluate prototype associations with hospital discharge diagnoses
• Validate interpretable digital phenotyping from physiological data

Innovation

• Prototype-based neural networks for ECG classification
• Unmodified inference on external clinical database
• Prototypes capture clinically meaningful physiological signatures