Hypertension Assessment via ECG and PPG Signals: An Evaluation Using MIMIC Database

Area of Science:

• Biomedical Engineering
• Cardiovascular Physiology
• Medical Diagnostics

Background:

• Cardiovascular diseases (CVDs) are a leading global health threat, significantly accelerated by hypertension.
• Hypertension, particularly prehypertension, often presents asymptomatically, leading to low awareness and control rates.
• Effective early detection and management are vital for preventing CVD complications.

Purpose of the Study:

• To investigate a novel hypertension management method using arterial wave propagation and photoplethysmography (PPG) morphological theories.
• To explore physiological changes across different blood pressure (BP) levels.
• To assess the efficacy of Pulse Arrival Time (PAT) and PPG features in classifying normotension, prehypertension, and hypertension.

Main Methods:

• Extracted Pulse Arrival Time (PAT) from electrocardiogram (ECG) and photoplethysmogram (PPG) signals.
• Extracted photoplethysmogram (PPG) morphological features.
• Classified BP categories using three feature sets: PAT only, PPG features only, and combined PAT and PPG features.

Main Results:

• Photoplethysmogram (PPG) features demonstrated higher accuracy in classifying BP categories compared to Pulse Arrival Time (PAT) alone.
• Combined PAT and PPG features significantly improved blood pressure classification performance.
• Achieved F1 scores of 84.34% (normotension vs. prehypertension), 94.84% (normotension vs. hypertension), and 88.49% (normotension + prehypertension vs. hypertension).

Conclusions:

• Simultaneous collection and analysis of ECG and PPG signals offer a promising approach for detecting hypertension.
• The integrated feature set provides a more robust method for BP classification.
• This non-invasive method holds potential for early hypertension screening and management.