Department of General Practice, Affiliated Hangzhou First People's Hospital, Zhejiang University School of Medicine, Hangzhou 310000, Zhejiang, China.
J Healthc Eng. 2021 Aug 9;2021:4845792. doi: 10.1155/2021/4845792. eCollection 2021.
The role of deep learning-based echocardiography in the diagnosis and evaluation of the effects of routine anti-heart-failure Western medicines was investigated in elderly patients with acute left heart failure (ALHF).
A total of 80 elderly patients with ALHF admitted to Affiliated Hangzhou First People's Hospital from August 2017 to February 2019 were selected as the research objects, and they were divided randomly into a control group and an observation group, with 40 cases in each group. Then, a deep convolutional neural network (DCNN) algorithm model was established, and image preprocessing was carried out. The binarized threshold segmentation was used for denoising, and the image was for illumination processing to balance the overall brightness of the image and increase the usable data of the model, so as to reduce the interference of subsequent feature extraction. Finally, the detailed module of deep convolutional layer network algorithm was realized. Besides, the patients from the control group were given routine echocardiography, and the observation group underwent echocardiography based on deep learning algorithm. Moreover, the hospitalization status of patients from the two groups was observed and recorded, including mortality rate, rehospitalization rate, average length of hospitalization, and hospitalization expenses. The diagnostic accuracy of the two examination methods was compared, and the electrocardiogram (ECG) and echocardiographic parameters as well as patients' quality of life were recorded in both groups at the basic state and 5 months after drug treatment.
After comparison, the rehospitalization rate and mortality rate of the observation group were lower than the rates of the control group, but the diagnostic accuracy was higher than that of the control group. However, the difference between the two groups of patients was not statistically marked ( > 0.05). The length and expenses of hospitalization of the observation group were both less than those of the control group. The specificity, sensitivity, and accuracy of the examination methods in the observation group were higher than those of the control group, and the differences were statistically marked ( < 0.05). There was a statistically great difference between the interventricular delay (IVD) of the echocardiographic parameters of patients from the two groups at the basic state and the left ventricular electromechanical delay (LVEMD) parameter values after 5 months of treatment ( < 0.05), but there was no significant difference in the other parameters. After treatment, the quality of life of patients from the two groups was improved, while the observation group was more marked than the control group ( < 0.05).
Echocardiography based on deep learning algorithm had high diagnostic accuracy and could reduce the possibility of cardiovascular events in patients with heart failure, so as to decrease the mortality rate and diagnosis and treatment costs. Moreover, it had an obvious diagnostic effect, which was conducive to the timely detection and treatment of clinical diseases.
探讨深度学习超声心动图在诊断和评估老年急性左心衰竭（ALHF）患者常规抗心力衰竭西药疗效中的作用。
选取 2017 年 8 月至 2019 年 2 月在杭州市第一人民医院附属老年 ALHF 患者 80 例为研究对象，随机分为对照组和观察组，各 40 例。然后建立深度卷积神经网络（DCNN）算法模型，并进行图像预处理。采用二值化阈值分割进行去噪，对图像进行光照处理，平衡图像整体亮度，增加模型可用数据，减少后续特征提取的干扰。最后，实现深度卷积层网络算法的详细模块。此外，对照组患者给予常规超声心动图检查，观察组患者给予基于深度学习算法的超声心动图检查。观察并记录两组患者的住院情况，包括死亡率、再住院率、平均住院时间和住院费用。比较两种检查方法的诊断准确性，并记录两组患者在基础状态和药物治疗后 5 个月的心电图（ECG）和超声心动图参数以及患者生活质量。
经比较，观察组再住院率和死亡率均低于对照组，但诊断准确率高于对照组。但两组患者差异无统计学意义（>0.05）。观察组患者的住院时间和费用均少于对照组。观察组检查方法的特异性、敏感性和准确性均高于对照组，差异有统计学意义（<0.05）。两组患者的超声心动图参数室间隔延迟（IVD）和左心室机电延迟（LVEMD）参数值在基础状态和治疗后 5 个月时差异均有统计学意义（<0.05），但其他参数差异无统计学意义。治疗后，两组患者的生活质量均有所改善，观察组改善情况优于对照组（<0.05）。
基于深度学习算法的超声心动图具有较高的诊断准确性，可降低心力衰竭患者心血管事件的发生概率，降低死亡率和诊断治疗费用。而且，它具有明显的诊断效果，有利于临床疾病的及时发现和治疗。
