What does this research mean for the field?

A deep learning model combining a multi-scale CNN and LSTM accurately estimates systolic blood pressure, diastolic blood pressure, and heart rate from ECG and PPG signals with mean absolute errors of 2.24 mmHg, 1.40 mmHg, and 0.84 bpm, respectively. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

January 1, 2022IEEE AccessOpen Access

Estimation of Beat-by-Beat Blood Pressure and Heart Rate From ECG and PPG Using a Fine-Tuned Deep CNN Model

Q: What is the clinical evidence from this study?

Study design: Other. Population: Blood pressure and heart rate estimation (n=1551). Intervention: Modified long-term recurrent convolutional network (multi-scale CNN and LSTM) vs. Other methods. Primary outcome: Mean absolute error (MAE) for predicting SBP, DBP, and HR.

View Full Paper

Ask AI

Bookmark

Key result

Deep learning accurately estimates blood pressure from ECG and PPG with ~2 mmHg systolic error.

n=1,551

Why the study?

Current cuffless blood pressure measurement technologies feature acceptable overall accuracy, motivating a sufficiently accurate cuffless estimation method using PPG and ECG signals.

Can a deep learning model accurately estimate beat-by-beat blood pressure and heart rate from single-channel ECG and PPG signals?

Population

1551 patients (12,000 records) from the PhysioNet MIMIC II dataset

Comparison

Deep learning model estimation vs invasive ABP signals

Design

Model development and validation study using ten-fold cross-validation

Authors

CYChih-Ta YenNational Taiwan University of Science and Technology SCSheng‐Nan ChangNational Taiwan University Hospital CLChenghong LiaoHainan University

Discussion

Member takes

Implication

May support cuffless monitoring development; leaves open prospective validation before clinical adoption.

Structured PICO

Can a deep learning model accurately estimate beat-by-beat blood pressure and heart rate from single-channel ECG and PPG signals?

Population

1551 patients from the PhysioNet MIMIC II dataset (12,000 records) with single-channel PPG, ECG (lead II), and invasive arterial BP signals.

Intervention

Deep learning model (modified long-term recurrent convolutional network comprising a multi-scale convolution network and a long short-term memory network) for cuffless estimation of BP and HR.

Comparator

Invasive arterial blood pressure (ABP) signals (ground truth).

Outcome

Mean absolute error (MAE) and standard deviation (SD) for predicting systolic blood pressure (SBP), diastolic blood pressure (DBP), and heart rate (HR).

A deep learning model using single-channel ECG and PPG signals can accurately estimate beat-by-beat blood pressure and heart rate, meeting established clinical standards.

Cite This Study

Yen et al. (2022) studied Blood pressure and heart rate estimation (n=1,551). Modified long-term recurrent convolutional network (multi-scale CNN and LSTM) vs. Other methods was evaluated on Mean absolute error (MAE) for predicting SBP, DBP, and HR. A deep learning model combining a multi-scale CNN and LSTM accurately estimated SBP, DBP, and HR from ECG and PPG signals with mean absolute errors of 2.24 mmHg, 1.40 mmHg, and 0.84 bpm.

synapsesocial.com/papers/6a14e4b60b551a5972391356 https://doi.org/10.1109/access.2022.3195857

View Full Paper

Ask AI

Bookmark

Estimation of Beat-by-Beat Blood Pressure and Heart Rate From ECG and PPG Using a Fine-Tuned Deep CNN Model

Discussion

Member takes

Implication

Structured PICO

Cite This Study

Estimation of Beat-by-Beat Blood Pressure and Heart Rate From ECG and PPG Using a Fine-Tuned Deep CNN Model

Discussion

Member takes

Implication

Structured PICO

Cite This Study

Also Consider

Also Consider