Blood pressure estimation based on pulse transit time with LANDMARC

Pulse transit time (PTT) has been proved to be an efficient indicator for blood pressure (BP) estimation recently. Most PTT-based approaches to BP estimation are mainly based on established models between BP and PTT. However, these approaches are usually unstable for different subjects due to various body conditions. In this paper, we propose a new method for PTT-based BP estimation by using a data-driven algorithm named LANDMARC. The proposed method was preliminarily tested on 10 subjects. Compared with reference BP recorded via a clinical sphygmomanometer, the experimental results indicated that the correlation coefficients for systolic BP (SBP) estimation and diastolic BP (DBP) estimation were 0.80 and 0.76, and the mean absolute differences were 5.11 mmHg and 4.74 mmHg, respectively.