The Estimation of Pressure of Blood Using ECG Image based on Artificial Intelligent Strategy

The monitoring of blood pressure is a main task in improving diagnosis and treatment of hypertension. Currently, the monitoring of blood pressure apply uncomfortable and unreliable cuff-based device. This cuff-based method cannot give an accurate measurement further the accurate intra-arterial pressure monitoring is not suitable for long-term monitoring, it left discomfort and not suitable for all patient. In order to handle these problems, the new non-invasive blood pressure measurement proposed. The proposed method used ECG images and deep learning model to predict the blood pressure of patients. The ECG images are generated from printing the ECG signal of the machine. This dataset is used to training the residual network (Resnet) model. The proposed method is able to predict the accurate blood pressure. The accuracy of the method is assessed using confusion matrix and ROC Curve. The accuracy of the proposed method is 98.5% and the ROC curve is relied above diagonal line. This new method can handle the non-accurate and discomfort problem of previous blood pressure measurement methods. Through this new method, the patient can obtain the improvement diagnosis and treatment for long-term medication.