Affiliation:
1. Department of Physical Education, Ganzhou Teachers College, Ganzhou 341000, Jiangxi, P. R. China
2. Hepatobiliary Surgery, Ganzhou People’s Hospital, Ganzhou 341000, Jiangxi, P. R. China
Abstract
This work was to explore the application of deep learning (DL) in identifying the neural mechanism of cardiovascular disease (CVD) and the role of physical exercise in the prevention and treatment of CVD. 200 cases of outpatient treatment in the hospital from January to December in 2021 were included as the research objects. 100 people with fitness exercise habits were sorted into the experiment group, and the other 100 cases without fitness exercise habit were in the control group. In addition, a DL-based CVD recognition model was constructed. The results showed that the detection effect of the back propagation (BP) algorithm under DL was better, with an average of over 99%. Heart rate variability (HRV) time domain analysis results showed that the Rrmaen, standard deviation of N-N interval (SDNN), and root mean square of the difference (RMSSD) of the experiment group were [Formula: see text][Formula: see text]ms, [Formula: see text][Formula: see text]ms, and [Formula: see text][Formula: see text]ms, respectively. These were observably higher than those of the control group ([Formula: see text]). In the HRV frequency domain analysis, the total frequency (TF) in the experiment group was [Formula: see text][Formula: see text]MS2, which was notably higher than that in the control group ([Formula: see text][Formula: see text]MS2, [Formula: see text]). The scores of anxiety and depression in the experiment group before exercise intervention were [Formula: see text] and [Formula: see text], respectively, which were highly decreased after intervention ([Formula: see text]). The CVD recognition model based on a DL algorithm could effectively identify CVD. Long-term regular exercise can effectively change the regulatory function of cardiovascular autonomic nerves and depression and anxiety states, which had popularization value.
Publisher
World Scientific Pub Co Pte Ltd
Cited by
2 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献