From Elastic to Viscoelastic: An EEMD-Enhanced Pulse Transit Time Model for Robust Blood Pressure Estimation
Achieves medical-grade RMSE of 5.22 mmHg systolic using EEMD-based damping metric
A new research paper from Boyuan Gu, Yijin Yang, Shuaiqi Cheng, and Xiaorong Ding introduces a method to improve cuffless blood pressure (BP) estimation using pulse transit time (PTT). Traditional PTT models rely on the Moens-Korteweg equation, which assumes arterial walls are purely elastic — a simplification that fails during rapid hemodynamic changes. The team addresses this by incorporating a viscoelastic compensation mechanism. They first reconstruct raw photoplethysmogram (PPG) signals using Modified Akima interpolation, then apply a robust Intersecting Tangent Method for precise pulse foot localization. Crucially, they use EEMD to isolate high-frequency Intrinsic Mode Functions, defining a novel "Viscoelastic Velocity Metric" that quantifies vascular damping (η · ε̇) typically ignored.
The framework was validated on a challenging subset of the MIMIC-II database (364 subjects, 28,525 cardiac cycles) with a high prevalence of hypertension (23.4%). Results show medical-grade accuracy: root mean square error of 5.22 mmHg for systolic and 3.65 mmHg for diastolic BP, with Pearson correlation coefficients above 0.97. These findings confirm that incorporating viscoelastic features significantly enhances robustness against vascular hysteresis, paving the way for more reliable continuous health monitoring without traditional cuffs.
- Uses EEMD (Ensemble Empirical Mode Decomposition) to extract a viscoelastic damping metric from PPG signals
- Achieves RMSE of 5.22 mmHg systolic and 3.65 mmHg diastolic on 28,525 cardiac cycles from 364 subjects
- Addresses limitation of elastic PTT models by modeling arterial viscoelasticity for more accurate BP estimation
Why It Matters
Enables more reliable cuffless blood pressure monitoring for continuous health tracking, especially in hypertensive patients.