An AIoT Wearable ECG Patch with Decision Tree for Arrhythmia Analysis

Abstract

This paper proposes a novel electrocardiogram (ECG) monitoring system with an artificial intelligence of things (AIoT) design and is based on a decision tree [1]. The proposed system includes a front-end device and a software system. The front-end device includes a solar charging circuit, a wireless charging circuit, and an analog front-end circuit. The analog front-end block is composed of an ECG sensing circuit, a microcontroller unit (MCU), and a Bluetooth low-energy (BLE) module. This paper also proposes an iOS-based app with decision-tree model. Real-time ECG data are classified by the decision-tree model into different arrhythmias, and the average delay latency is 500 ms. Meanwhile, with 4G or Wi-Fi, the collected ECG data are uploaded to the cloud server for training the decision tree, and the coefficients of the pre-trained decision tree will be sent back to update the decision-tree model in the app. An accuracy of 98.22% is achieved. Through the proposed AIoT system, doctors and users can perform long-term ECG monitoring, which is valuable for cardiovascular-disease diagnosis. Doctors can also instantly assist users with the web-user interface to meet the needs of telemedicine. The proposed AIoT system is also used in human trials in the National Cheng Kung University Hospital. The power consumption of the proposed front-end device is 8.25 mW, and it can be continuously used for up to 32 h with a 120 mAh lithium-ion battery. When operated on solar charging, the device can continue working until the solar cell runs out of power.