ECG diagnosis device based on machine learning

Abstract

ECG signal can reflect rich physiological information of human body. The health state of human body can be obtained through the analysis of ECG signal. At present, most ECG detectors can only detect ECG signal and calculate heart rate, but can not carry out intelligent diagnosis. A STM32 development board with a front-end acquisition module is used to collect the analog ECG signal generated by an ECG simulator in this works, and the collected signal is uploaded to the Edge Impulse platform for the construction and training of diagnostic model. Then, the trained ECG diagnosis neural network model is deployed in the main controller of the development board for heart rate calculation and ECG signal diagnosis. The device can not only monitor the user's ECG signal, but also display the graphical ECG signal, calculate heart rate value and classify diagnostic results.