A method of remote ECG monitoring, remote disease screening, and early-warning based on wavelet analysis, comprising the following steps:
- Step S1:
- collecting the ECG signal in real time after the user wears the wireless ECG signal acquisition device on the user's chest;
- Step S2:
- transmitting the collected ECG signal by an ECG signal acquisition patch in the wireless ECG signal acquisition device through an analog circuit and a digital processing circuit containing a compression algorithm to a low-power Bluetooth transmission circuit;
- and transmitting the collected ECG signal by the low-power Bluetooth transmission circuit to the mobile terminal;
- Step S3:
- receiving the ECG signal by a low-power Bluetooth receiving circuit in the mobile terminal from the low-power Bluetooth transmission circuit;
- and transmitting the ECG signal to the wavelet analysis algorithm module for analysis and processing;
- Step S4:
- processing the received ECG signal by a wavelet analysis algorithm module using the wavelet analysis algorithm:
- detecting each peak point of the ECG signal;
- calculating the time of each peak interval to obtain waveform features of the ECG signal;
- transmitting data and the waveform features of the ECG signal by the wavelet analysis algorithm module to the application client module;
- Step S5:
- establishing the user's personal account by the application client module;
- controlling the display module through the application client module to display the data and the waveform of the ECG signal obtained by wavelet analysis;
- sending an alarm by the alarm module controlled by the application client module when the user's ECG signal is significantly abnormal;
- Step S6:
- uploading the processed ECG signal by the application client module to the cloud storage platform;
- aggregating and storing the user's personal information, the ECG signal, and the waveform features of the ECG signal obtained by analysis, by the cloud storage platform;
- classifying each ECG waveform by the cloud storage platform using the support vector machine based cardiac diagnosis algorithm and a heart rate classification model;
- wherein classifications that can be realized include atrial premature beat, atrial fibrillation, atrial premature beat, ventricular flutter, atrial flutter, and normal heart rate;
- Step S7:
- generating an analysis report by the cloud storage platform when an abnormal heart rate is found;
- transmitting an ECG signal waveform of the abnormal heart rate and a heart rate classification to the application client module;
- feeding back the ECG signal waveform of the abnormal heart rate and the heart rate classification to the user;
- and exporting the report by the user to directly show the report to the doctor;
- and
- Step S8:
- modifying the heart rate classification by the doctor on the cloud storage platform when a judgment of the heart rate classification model is wrong;
- memorizing the ECG data by the classification model;
- readjusting parameters of the heart rate classification model;
- and establishing a specific classification model for each user by the cloud platform.