HIPNO: Symmetry-Aware Physics-Informed Neural Operators for Noninvasive Hemodynamic Inference

This paper introduces HIPNO (Hemodynamic Inference via Physics-informed Neural Operators), a novel approach for recovering hemodynamic state from non-invasive signals to expand access to advanced monitoring in surgery and intensive care. The method addresses the challenge of scale symmetry in physics-informed hemodynamic inference. By leveraging physics-informed neural operators, HIPNO enables continuous hemodynamic monitoring without the risks associated with invasive measurement.