Cardiac auscultation is a cost-effective and noninvasive technique for cardiovascular disease detection. Recently, various studies have been underway for cardiac auscultation using deep learning, not doctors. When training a deep learning network, it is important to secure large amount of high-quality data. However, medical data are difficult to obtain, and in most cases the number of abnormal classes is insufficient. In this study, data augmentation is used to supplement the insufficient amount of data, and data generalization to generate data suitable for convolutional neural networks (CNN) is proposed. We demonstrate performance improvements by inputting them into the CNN. Our method achieves an overall performance of 96%, 81%, and 90% for sensitivity, specificity, and F1-score, respectively. Diagnostic accuracy was improved by 18% compared to when it was not used. Particularly, it showed excellent detection success rate for abnormal heart sounds. The proposed method is expected to be applied to an automatic diagnosis system to detect heart abnormalities and help prevent heart disease through early detection.

Federated learning (FL) that can train using machine learning methods without moving data have attracted interest owing to the focus on data privacy. Several FL platforms and frameworks are being developed with various open datasets. However, FL has not yet been fully utilized in real-world projects; instead, centralized ML models are still being used for AI. Researchers who develop these models should be able to easily and conveniently apply custom data and models developed in a centralized environment to FL environments, deploy and train multiple clients, and manage the lifecycle of the entire FL process. This study proposes FLScalize to enable AI researchers to conveniently apply their own custom data and models to FL environments that can occur in the real world and to deploy and manage the FL lifecycle. FLScalize can be used to simulate system heterogeneity and data heterogeneity, both of which are FL issues that occur in real FL environments. Furthermore, FLScalize provides a manager component that continuously manages the FL client and server required for real-world FL tasks and realizes an FL lifecycle management implementation that enables continuous integration, deployment, and training.

This paper presents an important advancement in heart activity monitoring, focusing on non-contact sensor data, which tend to be noisy due to interference, and the limitations of non-contact (untact) technology. A preprocessing filter and optimal classification model are proposed to improve the accuracy and reliability of heart rate data measured by a non-contact Doppler radar sensor, and the results are compared to those of a contact heart rate sensor (Holter monitor). The MIT-BIH Arrhythmia Database of PhysioNet are used for learning, and the results from the non-contact sensor and Holter monitor are compared for verification. To train the abnormal heartbeat waveform classification model, (1) an optimal heart rate data separation window size is selected through iterative model comparison and used for data separation, and (2) meaningful indicators of heart rate variability are selected; the data are transformed and applied as model characteristics. The non-contact sensor data are then applied to three filter algorithms, and the accuracy is assessed by comparison with the contact sensor data using the trained abnormal heartbeat waveform classification model. Learning is performed using 12 classification models, and the accuracies of the models are compared. This study demonstrates an effective new method of transfer learning for contact data abnormality detection.