PPG-based blood pressure estimation can benefit from scalable multi-scale fusion neural networks and multi-task learning. (September 2022)

• Record Type: Journal Article
• Authors: Hu, Qihan; Wang, Daomiao; Yang, Cuiwei
• Is Part Of: Biomedical signal processing and control. Volume 78(2022)
• Publication Date: 2022-09
• Subjects: Blood pressure estimation -- Convolutional neural network -- Multi-scale fusion -- Multi-task learning -- PPG; 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.2022.103891
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: In the proposed scalable MSF block, parallel convolution layers can help obtain features of different receptive fields and then they were integrated to form multi-scale representation by the attention mechanism. Depthwise separable convolution borrowed from the MobileNet was introduced to keep the balance between parameter number and computation complexity. The above two advantages not only improved the richness of extracted features but also ensured the computation latency in real application. Multi-task learning was introduced to help backbone network obtain more robust representation. Task-specific subnetworks can refine the feature from the shared backbone network into the features more suitable for SBP and DBP estimation task, respectively. Multi-scale fusion backbone network combined with the task-specific subnetworks can improve BP estimation accuracy. Abstract: The continuous measurement of blood pressure (BP) plays an important role in preventing cardiovascular diseases. However, common cuff-based devices with cumbersome operations are difficult to realize continuous BP monitoring. We propose a continuous BP estimation method based on the end-to-end multi-scale fusion and multi-task learning neural network (MSF-MTLNet). It can use the 8-second PPG segment and its derivatives as input and automatically learn their multi-scale features to improve BP accuracy. The proposed method implements feature extraction by adaptively fusing multiple convolution kernels with different sizes through the attention mechanism. Simultaneously, multi-task learning was introduced to discover task-relevant features. A total of 277, 600 segments from 1825 patients in the University of California, Irvine database are used for experiments. They are divided into the independent training set, validation set, and test set according to the ratio of 8:1:1 in an inter-subject manner. The mean and standard deviation of estimation error for systolic blood pressure and diastolic blood pressure are 0.97 ± 8.87 mmHg and 0.55 ± 4.23 mmHg, respectively, which outperforms the current cutting-edge method. Our approach almost meets the Association for the Advancement of Medical Instrumentation standard. Considering the British Hypertension Society standards, it achieves an 'A' grade for DBP estimation and a 'B' grade for SBP estimation. Experimental results show that MSF-MTLNet can obtain promising precision in cuffless and continuous BP estimation.