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

2026-01-01

arXiv.org

AI Summary

This work addresses the high computational cost of existing deep learning methods for arrhythmia detection, which hinders their deployment on resource-constrained edge medical devices. The authors propose a lightweight framework that constructs a hybrid feature representation by integrating time–frequency wavelet decomposition with graph-theoretic descriptors—such as PageRank centrality—to render high-dimensional ECG data linearly separable. Feature space optimization is achieved through mutual information and recursive feature elimination, enabling the use of an ultra-lightweight linear classifier. The resulting model occupies only 8.54 KB and achieves 98.44% accuracy on the MIT-BIH and INCART datasets, with a single inference latency of 0.46 microseconds and a per-beat processing time of 52 milliseconds—demonstrating an order-of-magnitude improvement in efficiency over models such as KD-Light.

Technology Category

• Machine Learning: Learning on the Edge & Model Compression
• Computer Vision: Learning & Optimization for CV
• Natural Language Processing: Learning & Optimization for NLP

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

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.

Problem

• arrhythmia detection
• resource-constrained devices
• ECG signals
• Internet of Medical Things
• computational overhead

Innovation

• hybrid feature engineering
• resource-efficient ECG classification
• wavelet decomposition
• graph-theoretic descriptors
• linearly separable representation