Physics informed neural network for assessing hemodynamics
Physics-informed neural networks enable real-time, non-invasive diagnosis of CAD and MVD by predicting hemodynamic characteristics from angiographic data, overcoming the limitations of invasive methods and providing timely, quantitative assessments.
WO2026136838A1PCT designated stage Publication Date: 2026-06-25ANGIOINSIGHT INC
Patent Information
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- ANGIOINSIGHT INC
- Filing Date
- 2025-12-19
- Publication Date
- 2026-06-25
Smart Images
Figure US2025060598_25062026_PF_FP_ABST
Abstract
Systems and methods utilizing neural networks and associated architecture for assessing blood flow and vascular anomalies are disclosed herein. Systems and methods may assess blood flow and / or predict vascular anomalies by deriving point cloud data from angiographic data to form three-dimensional (3D) datasets, and may further input derived data into a neural network such that the neural network outputs and provides visual results in a 3D space and / or map displaying one or more of velocity streamlines, pressure contours and shear stress distributions representative of a patient or subject vasculature.
Need to check novelty before this filing date? Find Prior Art