Cardiovascular disease has become one of the main diseases threatening human life and health. This disease is very common and troublesome, and the existing medical resources are scarce, so it is necessary to use a computer-aided automatic diagnosis to overcome these limitations. A computer-aided diagnostic system can automatically diagnose through an electrocardiogram (ECG) signal. This paper proposes a novel deep-learning method for ECG classification based on adversarial domain adaptation, which solves the problem of insufficient-labeled training samples, improves the phenomenon of different data distribution caused by individual differences, and enhances the classification accuracy of cross-domain ECG signals with different data distributions. The proposed method includes three modules: multi-scale feature extraction F, domain discrimination D, and classification C. The module F, constitutive of three different parallel convolution blocks, is constructed to increase the breadth of features extracted from this module. The module D is composed of three convolutional blocks and a fully connected layer, which is to solve the problem of low model layers and low-feature abstraction. In the module C, the time features and the deep-learning extraction features are concatenated on the fully connected layer to enhance feature diversity. The effectiveness of the proposed method is verified by experiments, and the classification accuracy of the experimental electrical signals reaches 92.3%.

Extreme learning machine based on local receptive fields (ELM-LRFs) is a very fast method that can be used for feature extraction and classification. Bidirectional long-short time memory network (BLSTM), a widely used type of recurrent neural network (RNN) architecture, has showed excellent performance in time series processing fields. In this paper, we combine the superiority of above algorithms and propose a fast and accurate hybrid deep learning model which is named DELM-LRF-BLSTM for ECG signal recognition. This model uses the segmented heartbeats as input and employs a deep ELMLRF to gain significant local spatial features. Then we fed the features to a three-layer BLSTM and it can extract temporal features for ECG signal recognition. The combination of ELM-LRF and BLSTM can not only consider the local information in a heartbeat, but also consider the long-distance dependence between heartbeats. Experimental results on MIT-BIH Arrhythmia dataset show that the proposed DELMLRF- BLSTM algorithm has high accuracy and sensitivity, up to 99.32% and 97.15% respectively, which verifies the effectiveness and feasibility of the model. Moreover, only 6.1 millisecond is needed for once heartbeat recognition operation. Due to it’s high performance and low computational complexity, the proposed algorithm is feasible for practical use.

The goal of Clinical Named Entity Recognition (CNER) is to identify clinical terms from medical records, which is of great importance for subsequent clinical research. Most of the current Chinese CNER models use a single set of features that do not consider the linguistic characteristics of the Chinese language, e.g., they do not use both word and character features, and they lack morphological information and specialized lexical information on Chinese characters in the medical field. We propose a RoBerta Glyce-Flat Lattice Transformer-CRF (RG-FLAT-CRF) model to address this problem. The model uses a convolutional neural network to discern the morphological information hidden in Chinese characters, and a pre-trained model to obtain vectors with medical features. The different vectors are stitched together to form a multi-feature vector. To use lexical information and avoid the problem of word separation errors, the model uses a lattice structure to add lexical information associated with each word, which can be used to avoid the problem of word separation errors. The RG-FLAT-CRF model scored 95.61%, 85.17%, and 91.2% for F1 on the CCKS 2017, 2019, and 2020 datasets, respectively. We used statistical tests to compare with other models. The results show that most p-values less than 0.05 are statistically significant.

The precise identification of arrhythmia is critical in electrocardiogram (ECG) research. Many automatic classification methods have been suggested so far. However, efficient and accurate classification is still a challenge due to the limited feature extraction and model generalization ability. We integrate attention mechanism and residual skip connection into the U-Net (RA-UNET); besides, a skip connection between the RA-UNET and a residual block is executed as a residual attention convolutional neural network (RA-CNN) for accurate classification. The model was evaluated using the MIT-BIH arrhythmia database and achieved an accuracy of 98.5% and scores for the classes and of 82.8% and 91.7%, respectively, which is far superior to other approaches.

Remote photoplethysmography (rPPG) has recently attracted much attention due to its non-contact measurement convenience and great potential in health care and computer vision applications. Early rPPG studies were mostly developed on self-collected uncompressed video data, which limited their application in scenarios that require long-distance real-time video transmission, and also hindered the generation of large-scale publicly available benchmark datasets. In recent years, with the popularization of high-definition video and the rise of telemedicine, the pressure of storage and real-time video transmission under limited bandwidth have made the compression of rPPG video inevitable. However, video compression can adversely affect rPPG measurements. This is due to the fact that conventional video compression algorithms are not specifically proposed to preserve physiological signals. Based on this, we propose a video compression scheme specifically designed for rPPG application. The proposed approach consists of three main strategies: 1) facial ROI-based computational resource reallocation; 2) rPPG signal preserving bit resource reallocation; and 3) temporal domain up- and down-sampling coding. UBFC-rPPG, ECG-Fitness, and a self-collected dataset are used to evaluate the performance of the proposed method. The results demonstrate that the proposed method can preserve almost all physiological information after compressing the original video to 1/60 of its original size. The proposed method is expected to promote the development of telemedicine and deep learning techniques relying on large-scale datasets in the field of rPPG measurement.