Towards continuous blood pressure monitoring using speckle plethysmography

Ten people wore a contact speckle plethysmography (SPG) sensor alongside ECG, PPG and a reference blood pressure monitor. Pulse arrival time and waveform features from SPG correlated strongly with blood pressure, and the relationship varied less between people than it did for PPG, which suggests SPG might support a more general, less calibration-heavy BP model.

• Form factor: finger

Speckle Plethysmography (SPG) is recently gaining interest as an addition to Photoplethysmography (PPG), due to higher Signal-to-Noise Ratio and the reduced sensitivity to both ambient light changes and variation in skin melanin content. SPG can be obtained by analyzing the spatial changes in the speckle pattern generated by a coherent light source over time. Continuous Blood Pressure (BP) monitoring is a particularly relevant topic in the biomedical domain as elevated BP is a major risk factor for cardiovascular diseases. Noninvasive methodologies are being researched as an alternative to cuff-based methods, which are considered uncomfortable and cumbersome. Typical approaches are Pulse Arrival Time (PAT), using Electrocardiogram (ECG) signals and PPG, or Pulse Waveform Analysis (PWA), by analyzing PPG’s morphological features. These approaches, however, still lack accuracy and require personalized BP models. SPG could offer better performance for this type of application. Therefore, in this work, ten subjects were monitored using a contact SPG system synchronized with simultaneous reference recordings of ECG, PPG, and BP signals. PAT and PWA features were calculated from SPG and ECG. The same features were calculated from PPG, which were used for benchmarking purposes. Linear regression analysis was performed between the extracted features and the reference BP. In addition, the inter-subject variability in the feature BP-slope and the correlation coefficients were analyzed. The results show a high correlation between the SPG features and BP. Furthermore, the SPG features resulted in a reduced inter-subject variability compared to the PPG-derived features, possibly indicating that SPG could be used for a generalized BP model.

Verbatim from Semantic Scholar (opens in a new tab) · Abstract via Semantic Scholar; © the authors/publisher

No results from this paper on ChatPPG leaderboards yet. Browse the leaderboards.

Wavelengths

laser (λ not in abstract)