Automatic staging model of heart failure based on deep learning. (July 2019)

• Record Type: Journal Article
• Authors: Li, Dengao; Li, Xuemei; Zhao, Jumin; Bai, Xiaohong
• Is Part Of: Biomedical signal processing and control. Volume 52(2019)
• Page Start: 77
• Page End: 83
• Publication Date: 2019-07
• Subjects: Heart failure -- Staging model -- Deep learning -- Deep CNN-RNN model; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2019.03.009
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: Feature extraction, selection and classification procedures are combined in a single deep CNN-RNN structure. Denoising is not required. Ten-fold cross validation ensures the results are reliable and robust. Abstract: Heart failure (HF) is a disease that is harmful to human health. Recent advances in machine learning yielded new techniques to train deep neural networks, which resulted in highly successful applications in many pattern recognition tasks such as object detection and speech recognition. To improve the diagnostic accuracy of HF staging, this study evaluates the performance of deep learning-based models on combined features for its categorization. We proposed a novel deep convolutional neural network-Recurrent neural network (CNN-RNN) model for automatic staging of heart failure diseases in real-time and dynamically. We employed the data segmentation and data augmentation pre-processing dataset to make the classification performance of the proposed architecture better. Specifically, this paper use convolutional neural network (CNN) as a feature extractor instead of training the entire network to extract the characteristics of the electrocardiogram (ECG) signals and form a feature set. We combine the above feature set with other clinical features, feed the combined features to RNN for classification, and finally obtain 5 classification results. Experiments shows that the CNN-RNN model proposed in this paper achieved an accuracy of 97.6%, the sensitivity of 96.3%, specificity of 97.4% and proportion of 97.1% for two seconds of ECG segments. We obtained an accuracy, sensitivity, specificity and proportion of 96.2%, 96.9%, 95.7%, and 94.3% respectively for five seconds of ECG duration. The model can be used as an aid to help clinicians confirm their diagnosis.