Interpretable Rule Mining for Real-Time ECG Anomaly Detection in IoT Edge Sensors

Abstract

Electrocardiogram (ECG) analysis is widely used in the diagnosis of cardiovascular diseases. This paper proposes an explainable rule-mining strategy for prioritizing abnormal class detection in ECG data. The proposed method utilizes a biasedtrained Artificial Neural Network (ANN) with input features derived from an ECG beat sequence and formulates a set of rules at each node of an on-demand tree-like search algorithm. The rule base at each node is derived from a linear combination of the most impactful features identified using gradient analysis in an ANN. The final derived model is an explainable rule-based system that detects abnormal heartbeats based on statistical and morphological features from ECG. The model achieves the target sensitivity, and accuracy with a low run-time complexity through a comprehensive offline rule mining process and is trained using the MIT-BIH Arrhythmia Database. The system achieves an accuracy of 93% with only nine nodes and a test sensitivity of 90% and 80% respectively for VEB and SVEB beat types, when tested on previously unseen ECG data from the INCART database. The model performance and complexity can be easily adjusted based on the real-time resource constraints of a wearable sensor. The model was deployed on an ARM Cortex M4-based embedded device and is shown to achieve a > 50% reduction in sensor power consumption when only abnormal beats are wirelessly transmitted. i.e RF transmission is gated using the model output and transmission is disabled when the subject's ECG is normal. The proposed technique is highly suited for healthcare applications because of its explainability, lower complexity, and real-time flexibility when deployed in the Internet of Things (IoT) enabled wearable edge sensors.

AI-Driven Real-Time Classification of ECG Signals for Cardiac Monitoring Using i-AlexNet Architecture

“…It was necessary to create a multiclass modelling categorization in order to make a precise determination. The feature identification method established by investigators in [29] was through the use of ECG, and the feature subset was chosen using kernel-based complicated coarse groups. Subsequently, optimization techniques satisfying several objectives were used to generate the classification of arrhythmia based on electrocardiogram (MC-ECG) for different varieties of labels.…”

A Systematic Review of Machine Learning and IoT Applied to the Prediction and Monitoring of Cardiovascular Diseases

“…The study of Yu et al [ 118 ] was the only proposal to classify stroke, while only Rodriguez et al [ 117 ] classified saturated oxygen and Sahani et al [ 119 ] focused on carotid disease. Ying et al [ 116 ], Sivapalan et al [ 121 ], and Rahman et al [ 120 ] focused on ECG abnormalities and ECG noise.…”

“…Appl 2 [ 39 , 40 , 83 , 90 ] Future Gener Comput Syst. 4 [ 196 ] Health Technol 1 [ 61 ] ICT Express 1 [ 37 , 43 , 45 , 49 , 52 , 54 , 55 , 56 , 80 , 84 , 94 , 115 , 133 , 158 , 169 , 170 , 191 , 199 , 205 ] IEEE Access 18 [ 77 , 106 , 107 , 110 , 121 ] IEEE Internet Things J. 5 [ 66 , …”

“…Res. Public Health 1 [ 50 , 77 , 106 , 107 , 110 , 121 ] Internet Things 6 [ 74 ] J. Am Heart Assoc 1 [ 192 ] J. Cardiovasc.…”