Smartphone-Based Blood Pressure Measurement Using Transdermal Optical Imaging Technology.

Area of Science:

• Biomedical Engineering
• Medical Technology
• Artificial Intelligence in Healthcare

Background:

• Traditional cuff-based blood pressure measurement methods are often uncomfortable and inconvenient for users.
• A novel smartphone-based technology, transdermal optical imaging, offers a contactless approach to blood pressure assessment.
• This technology analyzes subtle facial blood flow variations captured via smartphone video.

Purpose of the Study:

• To investigate the feasibility of determining blood pressure using contactless transdermal optical imaging.
• To develop and validate machine learning models for predicting blood pressure from facial blood flow signals.
• To assess the accuracy of smartphone-based blood pressure measurement in normotensive adults.

Main Methods:

• 1328 normotensive adults were enrolled in the study.
• Advanced machine learning algorithms were employed to create predictive computational models.
• The dataset was divided into training (70%), testing (15%), and validation (15%) subsets for model development and performance evaluation.

Main Results:

• The developed models demonstrated high accuracy in predicting blood pressure.
• Systolic pressure prediction bias±SD was 0.39±7.30 mm Hg.
• Diastolic and pulse pressure prediction bias±SD were -0.20±6.00 mm Hg and 0.52±6.42 mm Hg, respectively.

Conclusions:

• The study's findings in normotensive adults show results within 5±8 mm Hg of reference measurements.
• Further research is required to validate model performance against international accuracy standards across a wider range of blood pressures.
• Contactless smartphone-based blood pressure monitoring shows promising potential for convenient and comfortable health assessment.