Study On The Portable Mobile Medical ECG Detection And Diagnosis System

In recent years, wearable device has become research hotspot. It shows the research value and applied potential in many fields. With the improvement of living standards, people are more concerned about health issues and their disease prevention awareness is also increasing. But the number of deaths caused by cardiovascular diseases is still high in our country. There are still risks and challenges in the medical conditions of this aspect at present.In order to take preventive measures to detect heart disease and provide sufficient treatment time, it has become a trend that medical treatment of hospitals transform into health care prevention in the future medical direction of our country. Miniaturization, intelligent, portable medical monitoring equipment meet the requirements, it is possible into the community, into the home to thousands of families to provide detection and monitoring. Wearable ECG device can get ECG signal of the body for a long time monitoring and diagnostic ECG information to detect heart problems. Wearable ECG equipment has begun to come into our lives, but the market isn't mature. From the foundation of the ECG monitoring, we propose a wearable ECG detection and diagnosis system. The paper is mainly from the wearable ECG hardware design, ECG filtering and ECG feature detect and diagnosis three aspects to elaborate.Firstly, according to the principles of wearable ECG system the paper analyzes and designs the wearable ECG hardware detection system from the ECG acquisition circuit,Bluetooth communication circuit, power supply circuit and MCU control circuit.Secondly, since the ECG acquisition process is easy to couple into interfering signals of the respective frequency band, it needs filter out interference noise form ECG signal. Based on Integer IIR filter design principle and the idea of the wavelet filter, the paper designed an improved integer IIR digital filter to filter out baseline wander, 50 Hz and high frequency noise. It achieved good filtering performance.Finally, the paper compared the difference threshold algorithm and Mallat wavelet transform algorithm on ECG feature detection and chose the higher accuracy and the better stability Mallat wavelet transform algorithm for ECG feature detection. Then, based on the basis of a single BP neural network, the paper designed a double BP neural network algorithmto diagnosis premature atrial contraction and ventricular premature contraction. The recognition rate can reach 97%.Wearable ECG detection and diagnosis system designed in this paper can acquire and diagnose ECG information of human body. It achieved a wearable design and proposed an improved integer IIR filter algorithm and a double BP neural network algorithm. Based on the validation of the algorithm using the MIT/BIH ECG database, it can acquire and analyze the ECG data of subjects and obtain an accurate diagnosis result.