Generalization Studies of Neural Network Models for Cardiac Disease Detection Using Limited Channel ECG
This work addresses generalization challenges in machine learning for cardiac disease detection, which is incremental as it builds on existing deep learning methods by incorporating unsupervised techniques.
The paper tackled the problem of detecting cardiovascular conditions from limited-channel ECG measurements by proposing an unsupervised learning approach to improve generalization to unseen classes, achieving significant improvements in F1-scores.
Acceleration of machine learning research in healthcare is challenged by lack of large annotated and balanced datasets. Furthermore, dealing with measurement inaccuracies and exploiting unsupervised data are considered to be central to improving existing solutions. In particular, a primary objective in predictive modeling is to generalize well to both unseen variations within the observed classes, and unseen classes. In this work, we consider such a challenging problem in machine learning driven diagnosis: detecting a gamut of cardiovascular conditions (e.g. infarction, dysrhythmia etc.) from limited channel ECG measurements. Though deep neural networks have achieved unprecedented success in predictive modeling, they rely solely on discriminative models that can generalize poorly to unseen classes. We argue that unsupervised learning can be utilized to construct effective latent spaces that facilitate better generalization. This work extensively compares the generalization of our proposed approach against a state-of-the-art deep learning solution. Our results show significant improvements in F1-scores.