Attractor-Vascular Coupling Theory Enables AAMI-Standard Cuffless Blood Pressure Estimation via Smartphone PPG
Researchers Timothy Oladunni and Farouk Ganiyu Adewumi propose the Attractor-Vascular Coupling Theory (AVCT), a mathematical framework demonstrating that cardiac attractor geometry encodes sufficient blood pressure information for clinical-grade estimation. Published on arXiv, the study validates AVCT using a calibrated cuffless model based on photoplethysmography (PPG) data. The model, utilizing LightGBM and features like Pulse Transit Time and Cardiac Stability Index, was evaluated on 46 subjects from BIDMC ICU and VitalDB datasets. It achieved mean absolute errors of 2.05 mmHg for systolic and 1.67 mmHg for diastolic blood pressure, satisfying AAMI/IEEE SP10 standards. Notably, a smartphone-only PPG ablation matched the performance of combined ECG+PPG models, proving that accurate blood pressure tracking is feasible using only a smartphone camera. The theory reduces estimation error by 91.5% through calibration and offers explainable AI features grounded in nonlinear dynamical systems, surpassing prior generalized results with fewer sensors.
Wire timeline
Attractor-Vascular Coupling Theory Enables AAMI-Standard Cuffless Blood Pressure Estimation via Smartphone PPG
Researchers Timothy Oladunni and Farouk Ganiyu Adewumi propose the Attractor-Vascular Coupling Theory (AVCT), a mathematical framework demonstrating that cardiac attractor geometry encodes sufficient blood pressure information for clinical-grade estimation. Published on arXiv, the study validates AVCT using a calibrated cuffless model based on photoplethysmography (PPG) data. The model, utilizing LightGBM and features like Pulse Transit Time and Cardiac Stability Index, was evaluated on 46 subjects from BIDMC ICU and VitalDB datasets. It achieved mean absolute errors of 2.05 mmHg for systolic and 1.67 mmHg for diastolic blood pressure, satisfying AAMI/IEEE SP10 standards. Notably, a smartphone-only PPG ablation matched the performance of combined ECG+PPG models, proving that accurate blood pressure tracking is feasible using only a smartphone camera. The theory reduces estimation error by 91.5% through calibration and offers explainable AI features grounded in nonlinear dynamical systems, surpassing prior generalized results with fewer sensors.
cs.AI updates on arXiv.org