SUNQUI: Ambulatory Arrhythmia Monitoring Device Based on Artificial Intelligence Pre-Diagnosis in Electrocardiography

This study proposes a portable electrocardiographic (ECG) monitoring device for real-time arrhythmia detection, addressing limitations in accessibility and infrastructure in resource limited areas. The device uses a modified one-dimensional convolutional neural network (CNN) based on the AlexNet architecture to classify heart rhythms, including sinus rhythm, atrial fibrillation, and bradycardia. It integrates an AD8232 ECG sensor, ESP32 microcontroller, and wireless communication module, providing continuous ECG data collection and real-time analysis. Data is transmitted to a desktop platform for remote monitoring by healthcare professionals. The device was tested using patient data from PhysioNet, achieving 97% accuracy, 97.02% sensitivity, and 99.06% specificity, demonstrating its effectiveness in arrhythmia detection.Clinical relevance- This device provides a cost-effective, portable solution for continuous ECG monitoring, enabling real-time arrhythmia detection in settings without access to cardiology specialists. It could offer clinicians the ability to remotely monitor patients for critical conditions such as atrial fibrillation and bradycardia, facilitating timely intervention and improving patient outcomes. However, real-time monitoring may be affected by artifacts and noise, which need to be addressed and validated in clinical trials before drawing conclusions about clinical impact.