Prototype Learning to Create Refined Interpretable Digital Phenotypes from ECGs

TL;DR

This study demonstrates that prototype-based neural networks trained on ECG data can produce interpretable, clinically meaningful phenotypes that generalize across datasets and relate to diverse health conditions.

Contribution

The paper shows that prototypes from ECG classification models are associated with broad clinical phenotypes and outperform traditional class predictions in capturing meaningful physiological variations.

Findings

Prototypes strongly associate with hospital discharge diagnoses.

Models achieve high AUCs for cardiac and non-cardiac conditions.

Prototypes reveal intra-class physiological diversity.

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.