Optimized Hybrid Feature Engineering for Resource-Efficient Arrhythmia Detection in ECG Signals: An Optimization Framework

TL;DR

This paper introduces a resource-efficient feature engineering framework for arrhythmia detection in ECG signals, achieving high accuracy with minimal computational resources suitable for edge devices.

Contribution

The study presents a hybrid feature engineering approach combining wavelet and graph-theoretic descriptors, optimized for ultra-lightweight, interpretable classifiers in resource-constrained settings.

Findings

98.44% diagnostic accuracy on ECG datasets

Model footprint of only 8.54 KB

Real-time inference with 0.46 μs latency

Abstract

Cardiovascular diseases, particularly arrhythmias, remain a leading global cause of mortality, necessitating continuous monitoring via the Internet of Medical Things (IoMT). However, state-of-the-art deep learning approaches often impose prohibitive computational overheads, rendering them unsuitable for resource-constrained edge devices. This study proposes a resource-efficient, data-centric framework that prioritizes feature engineering over complexity. Our optimized pipeline makes the complex, high-dimensional arrhythmia data linearly separable. This is achieved by integrating time-frequency wavelet decompositions with graph-theoretic structural descriptors, such as PageRank centrality. This hybrid feature space, combining wavelet decompositions and graph-theoretic descriptors, is then refined using mutual information and recursive elimination, enabling interpretable, ultra-lightweight linear classifiers. Validation on the MIT-BIH and INCART datasets yields 98.44% diagnostic accuracy with an 8.54 KB model footprint. The system achieves 0.46 $\mu$s classification inference latency within a 52 ms per-beat pipeline, ensuring real-time operation. These outcomes provide an order-of-magnitude efficiency gain over compressed models, such as KD-Light (25 KB, 96.32% accuracy), advancing battery-less cardiac sensors.