Hypertension (HPT) is a chronic disease characterized by the consistent elevation of arterial blood pressure, which is considered to be a significant risk factor for conditions such as stroke, coronary artery disease, and heart failure. The detection and continuous monitoring of HPT can be a demanding process. As a non-contact measuring method, the ballistocardiography (BCG) signal characterizes the repetitive body motion resulting from the forceful ejection of blood into the major blood vessels during each heartbeat. Therefore, it can be applied for HPT detection. HPT detection with BCG signals remains a challenging task. In this study, we propose an end-to-end deep convolutional model BH-Net for HPT detection through BCG signals. We also propose a data augmentation scheme by selecting the J-peak neighborhoods from the BCG time sequences for hypertension detection. Rigorously evaluated via a public data-set, we report an average accuracy of 97.93% and an average F1-score of 97.62%, outperforming the comparative state-of-the-art methods. We also report that the performance of the traditional machine learning methods and the comparative deep learning models was improved with the proposed data augmentation scheme.

Automatic classification of electrocardiogram (ECG) signals plays a crucial role in the early prevention and diagnosis of cardiovascular diseases. While ECG signals can be used for the diagnosis of various diseases, their pathological characteristics exhibit minimal variations, posing a challenge to automatic classification models. Existing methods primarily utilize convolutional neural networks to extract ECG signal features for classification, which may not fully capture the pathological feature differences of different diseases. Transformer networks have advantages in feature extraction for sequence data, but the complete network is complex and relies on large-scale datasets. To address these challenges, we propose a single-layer Transformer network called Multi-Scale Shifted Windows Transformer Networks (MSW-Transformer), which uses a multi-window sliding attention mechanism at different scales to capture features in different dimensions. The self-attention is restricted to non-overlapping local windows via shifted windows, and different window scales have different receptive fields. A learnable feature fusion method is then proposed to integrate features from different windows to further enhance model performance. Furthermore, we visualize the attention mechanism of the multi-window shifted mechanism to achieve better clinical interpretation in the ECG classification task. The proposed model achieves state-of-the-art performance on five classification tasks of the PTBXL-2020 12-lead ECG dataset, which includes 5 diagnostic superclasses, 23 diagnostic subclasses, 12 rhythm classes, 17 morphology classes, and 44 diagnosis classes, with average macro-F1 scores of 77.85%, 47.57%, 66.13%, 34.60%, and 34.29%, and average sample-F1 scores of 81.26%, 68.27%, 91.32%, 50.07%, and 63.19%, respectively.