Continuous noninvasive blood pressure estimation using tissue blood flow measured by diffuse correlation...

Area of Science:

• Biomedical Engineering
• Medical Devices
• Cardiovascular Health

Background:

• Continuous blood pressure (BP) monitoring is crucial for cardiovascular health management.
• Current gold standard for continuous BP measurement is invasive (catheter).
• Diffuse Correlation Spectroscopy (DCS) measures tissue blood flow (BF) noninvasively.

Purpose of the Study:

• To develop and evaluate a deep learning model (BFBP) for continuous BP estimation using noninvasive tissue BF measurements from DCS.
• To assess the feasibility and accuracy of the BFBP model for real-time BP monitoring.

Main Methods:

• Developed a DCS device for measuring tissue BF.
• Utilized a deep learning approach (BFBP model) to correlate tissue BF with BP.
• Conducted in vivo experiments on 12 subjects for data collection.
• Validated the model against an open BP dataset (UC Irvine database).

Main Results:

• The BFBP model achieved mean absolute errors of 5.54±5.03 mm Hg for systolic BP and 1.71±2.86 mm Hg for diastolic BP.
• Continuous BP estimates achieved Grade A by the British Hypertension Society standard.
• Demonstrated feasible continuous BP estimation using noninvasive tissue BF.

Conclusions:

• DCS combined with deep learning offers a novel, noninvasive approach for continuous BP monitoring.
• The BFBP model shows high accuracy and meets clinical standards for BP estimation.
• This technology has the potential to improve cardiovascular health management through continuous, noninvasive BP tracking.