Medical Robot-Assisted Electrocardiogram Real-Time Monitoring and Arrhythmia Diagnosis System

A medical robot-assisted electrocardiogram (ECG) real-time monitoring and arrhythmia diagnosis system was designed. The ECG signal acquisition front-end utilized the AD8232 bioelectricity acquisition chip and embedded a digital filtering algorithm in the program. The front-end node transmits the collected ECG data to the QT host computer, where a two-dimensional convolutional neural network classifier is deployed to automatically compare the abnormal ECG data and obtain the classification results of arrhythmia data. The classification results are displayed in the host computer interface alongside the real-time ECG waveform, human heart rate, and human body temperature. The test results show that the accuracy of resting heart rate measurement is maintained within ±2%, the recall rate for seven types of heartbeats is higher than 90%, the average sensitivity is 98.59%, and the average specificity is 99.08%.