Future possibilities for artificial intelligence in the practical management of hypertension.
Journal:
Hypertension research : official journal of the Japanese Society of Hypertension
PMID:
32655135
Abstract
The use of artificial intelligence in numerous prediction and classification tasks, including clinical research and healthcare management, is becoming increasingly more common. This review describes the current status and a future possibility for artificial intelligence in blood pressure management, that is, the possibility of accurately predicting and estimating blood pressure using large-scale data, such as personal health records and electronic medical records. Individual blood pressure continuously changes because of lifestyle habits and the environment. This review focuses on two topics regarding controlling changing blood pressure: a novel blood pressure measurement system and blood pressure analysis using artificial intelligence. Regarding the novel blood pressure measurement system, we compare the conventional cuff-less method with the analysis of pulse waves using artificial intelligence for blood pressure estimation. Then, we describe the prediction of future blood pressure values using machine learning and deep learning. In addition, we summarize factor analysis using "explainable AI" to solve a black-box problem of artificial intelligence. Overall, we show that artificial intelligence is advantageous for hypertension management and can be used to establish clinical evidence for the practical management of hypertension.