FMM-Head: Enhancing Autoencoder-based ECG anomaly detection with prior knowledge

TL;DR

This paper introduces FMM-Head, an ECG anomaly detection model that incorporates prior ECG knowledge into an autoencoder, achieving higher accuracy, smaller size, and faster processing than existing methods.

Contribution

The paper proposes replacing the decoder in an autoencoder with a prior knowledge-based reconstruction head, improving ECG anomaly detection with explainability and efficiency.

Findings

Achieves up to 0.31 higher AUROC than state-of-the-art models.

Reduces model size by half compared to original autoencoders.

Enables real-time ECG anomaly detection with significantly lower processing time.

Abstract

Detecting anomalies in electrocardiogram data is crucial to identifying deviations from normal heartbeat patterns and providing timely intervention to at-risk patients. Various AutoEncoder models (AE) have been proposed to tackle the anomaly detection task with ML. However, these models do not consider the specific patterns of ECG leads and are unexplainable black boxes. In contrast, we replace the decoding part of the AE with a reconstruction head (namely, FMM-Head) based on prior knowledge of the ECG shape. Our model consistently achieves higher anomaly detection capabilities than state-of-the-art models, up to 0.31 increase in area under the ROC curve (AUROC), with as little as half the original model size and explainable extracted features. The processing time of our model is four orders of magnitude lower than solving an optimization problem to obtain the same parameters, thus making it suitable for real-time ECG parameters extraction and anomaly detection.