Abstract:Continuous blood pressure (BP) monitoring is crucial for diagnosing and preventing cardiovascular disease (CVD). However, existing approaches for continuous cuffless BP monitoring using photoplethysmogram (PPG) and electrocardiogram (ECG) signals suffer from instability and susceptibility to various factors. This poses a challenge in accurately estimating BP levels, hindering effective disease diagnosis and prevention. Therefore, there is a need for an improved method that can provide reliable and accurate continuous BP estimation from PPG and ECG signals, overcoming the limitations of existing approaches. In this study, we proposed a deep regression model with state space reconstruction (SSR) for continuous BP estimation. A feature voting system with a variety of feature selection algorithms is introduced to select the optimal feature set of PPG and ECG signals. The SSR technique is applied to feature data to reveal useful hidden information. The proposed method is evaluated based on 660 subjects from a well-known benchmark dataset and a multi-day BP dataset. Random forest and the proposed deep regression model are tested to show the advantages of using SSR on feature data. The results show a promising performance of the improved deep regression model. On the benchmark dataset, the root mean square error (RMSE) and mean absolute error (MAE) of the improved deep regression model are 3.613 and 2.765 mmHg respectively for systolic BP (SBP), 1.978 and 1.543 mmHg for diastolic BP (DBP). The results achieved high accuracy for estimating SBP and DBP according to the British Hypertension Society (BHS) standard. On the multi-day BP dataset, the proposed model achieved RMSE of 5.387, 3.338 and 3.611 mmHg for SBP and achieved MAE of 4.115, 2.553, and 2.927 mmHg for DBP. Additionally, the robustness of our proposed model is validated by adding random noise into the PPG signals. The results demonstrate that the proposed deep regression model with SSR can improve the performance of BP estimation. It is possible to apply our proposed method further to develop a wearable device for real-time BP monitoring.

Continuous Blood Pressure Estimation Based on Two-Domain Fusion Model

Hybrid CNN-SVR Blood Pressure Estimation Model Using ECG and PPG signals

Continuous blood pressure measurement from one-channel electrocardiogram signal using deep-learning techniques

Non-invasive continuous blood pressure prediction based on ECG and PPG fusion map

Continuous Blood Pressure Estimation Based on Multi-Scale Feature Extraction by the Neural Network With Multi-Task Learning

A Multi-Parameter Fusion Method for Cuffless Continuous Blood Pressure Estimation Based on Electrocardiogram and Photoplethysmogram

Non-Invasive Continuous Blood-Pressure Monitoring Models Based on Photoplethysmography and Electrocardiography

PPG-based blood pressure estimation can benefit from scalable multi-scale fusion neural networks and multi-task learning

A Novel Method for Continuous Blood Pressure Estimation Based on a Single-Channel Photoplethysmogram Signal.

Blood pressure prediction based on multi-sensor information fusion of electrocardiogram, photoplethysmography, and pressure pulse waveform

BPNet: A multi-modal fusion neural network for blood pressure estimation using ECG and PPG

Scenario Adaptive Cuffless Blood Pressure Estimation by Integrating Cardiovascular Coupling Effects

Blood pressure monitoring via double sandwich-structured triboelectric sensors and deep learning models

Multi-Dimensional Feature Combination Method for Continuous Blood Pressure Measurement Based on Wrist PPG Sensor.

ACNN-BiLSTM: A Deep Learning Approach for Continuous Noninvasive Blood Pressure Measurement Using Multi-Wavelength PPG Fusion

A Novel Model for Continuous Cuff-less Blood Pressure Estimation.

A new deep learning framework based on blood pressure range constraint for continuous cuffless BP estimation

A mixed attention-gated U-Net for continuous cuffless blood pressure estimation

Hybrid Modeling on Reconstitution of Continuous Arterial Blood Pressure Using Finger Photoplethysmography.

An improved deep regression model with state space reconstruction for continuous blood pressure estimation

Key Feature Selection and Model Analysis for Blood Pressure Estimation from Electrocardiogram, Ballistocardiogram and Photoplethysmogram