Integrated thermal sensing of vascular morphology for haemodynamic monitoring

Wearable haemodynamic monitoring systems are limited by the lack of in situ measurements of vascular morphology, which introduces systematic errors in the multi-physics field conduction model across tissue layers. Here we present an integrated approach that combines an analytical model with a miniaturized sensor for non-invasive, high-precision characterization of vascular morphology. By analysing transient thermal responses at the skin surface during localized heating, we establish quantitative relationships between vascular depth (H_Tube), inner diameter (D_Tube) and the skin thermal relaxation time constant (τ). A 20 × 2 mm² sensor captures spatially heterogeneous temperature fields, enabling the decoupling of morphological parameters through the model. Measurements of human arm veins show accuracies of approximately 0.3 mm for H_Tube and 0.2 mm for D_Tube, comparable to those obtained using ultrasound imaging. This approach reduces modelling uncertainty in haemodynamic parameter estimation, offering a wearable-compatible strategy for real-time vascular health monitoring.