Hybrid 1D-CNN and SVM Approach for ECG Arrhythmia Classification-using Constrained Hardware

Cardiovascular Diseases (CVDs) remain a leading global health concern, necessitating accessible and accurate monitoring solutions. This paper presents a novel approach to real-time arrhythmia classification using a hybrid 1D Convolutional Neural Network (CNN) and Support Vector Machine (SVM) architecture. The system uses the AD8232 ECG sensor and Arduino Uno for continuous data acquisition at approximately 360 Hz. The 1D CNN extracts robust, 64-dimensional features from 300-sample segments, which are then fed into the computationally efficient SVM classifier. This integrated pipeline achieved a high accuracy of 99.16% on the test set, with an operational latency under 1.0 second per prediction, confirming its viability for resource-constrained, real-time monitoring devices.