Hypertension Screening via Awake-Sleep Differences in Photoplethysmogram Signals

BackgroundHypertension is a major risk factor for cardiovascular diseases. This study proposes a novel hypertension screening framework based on awake-sleep differences in photoplethysmography (PPG) indices, using machine learning. We hypothesised that normotensive individuals exhibit greater PPG variation between awake and sleep states than unmanaged hypertensive individuals. MethodsThe Aurora-BP dataset (n=180; 138 normotensive, 42 hypertensive) was used for model development, with 18 subjects reserved for internal testing. External validation was performed using the independent CUHK-BP dataset (n=26; 11 normotensive, 15 hypertensive). Twenty PPG-based indices were extracted, and subject-level p-values from Mann-Whitney U-tests comparing awake and sleep periods were used as model features. Discretised p-values served as inputs for four machine learning models. ResultsThe Support Vector Machine (SVM) achieved the highest performance, with 81.1 {+/-} 8.4% accuracy and 82.8 {+/-} 8.1% F1-score on the internal test set using all indices. On the external test set, the SVM using only temporal indices achieved 84.6% accuracy and 86.7% F1-score. Temporal indices, especially those linked to the dicrotic notch, showed strong generalisability across datasets. ConclusionThe study demonstrates the feasibility of awake-sleep PPG analysis for hypertension screening, highlighting the potential of wearable PPG devices for ambulatory monitoring. Novelty and RelevanceO_ST_ABSWhat Is New?C_ST_ABSO_LIThis study introduces a novel machine learning-based hypertension screening method using awake-sleep differences in ambulatory photoplethysmogram (PPG) signals, validated on both internal and external datasets with a novel p-value-based feature selection strategy. C_LI What Is Relevant?O_LIThe proposed approach enables non-invasive, cuff-free hypertension detection using wearable PPG data, offering improved generalisability and real-world applicability over traditional models. C_LI Clinical/Pathophysiological Implications?O_LIThe ability to detect altered nocturnal cardiovascular patterns from PPG may support early identification of unmanaged hypertension and promote more accessible long-term monitoring. C_LI