Abstract:INTRODUCTION: Many studies have explored the application of machine learning and neural networks in extracting critical diagnostic information from structural and functional medical imaging. While these methods show potential for improving efficiency, concerns arise when interpreting subtle imaging features, especially when training data is limited. Physics-informed neural networks (PINNs) address these issues by incorporating governing equations from physical and mechanical models into the analysis. METHODS: We examined the use of a PINN that enforces the convection equation, relating contrast media propagation to blood velocity, the Navier-Stokes equations for velocity and pressure distributions, and the conservation of mass requirement to calculate velocity and pressure distributions within patient-specific blood vessels. We also implemented a boundary condition that accounts for real contrast-media propagation from 1000 fps high-speed angiographic (HSA) image sequences, allowing for an assumption-free problem space within a medical imaging framework. Velocity fields and pressure gradients were calculated for flow in aneurysms and carotid bifurcations. RESULTS: Our method demonstrates comparable results to computational fluid dynamics (CFD) without requiring manual data processing, significantly improving the efficiency of calculating high-resolution velocity and pressure fields. By integrating AI with physics modeling, this novel approach holds the potential for advancing neurovascular diagnostics and providing more accurate, personalized treatment plans for patients with neurovascular pathologies. CONCLUSIONS: This innovative approach, integrating physics modeling and artificial intelligence, offers more accurate and personalized diagnostics for patients with neurovascular pathologies. The PINN method not only achieves results comparable to computational fluid dynamics without the need for manual data processing but also greatly enhances the efficiency of calculating high-resolution velocity and pressure fields.

Performing particle image velocimetry using artificial neural networks: a proof-of-concept

Particle image velocimetry based on a deep neural network

Dense Motion Estimation of Particle Images Via a Convolutional Neural Network

Filtering Enhanced Tomographic PIV Reconstruction Based on Deep Neural Networks

Particle image velocimetry analysis with simultaneous uncertainty quantification using Bayesian neural networks

Particle Identification in Particle Tracking Velocimetry Using Two-Stage Neural Networks

Particle Image Velocimetry Based on a Deep Learning Motion Estimator.

DeepPTV: Particle Tracking Velocimetry for Complex Flow Motion via Deep Neural Networks

Deep-PIV: a new framework of PIV using deep learning techniques

A Calibration-Informed Deep Learning Model for Three-Dimensional Particle Reconstruction of Volumetric Particle Image Velocimetry

A robust single-pixel particle image velocimetry based on fully convolutional networks with cross-correlation embedded

Demonstration of Vector Flow Imaging using Convolutional Neural Networks

Best of Both Worlds: Hybrid SNN-ANN Architecture for Event-based Optical Flow Estimation

Generalization of deep recurrent optical flow estimation for particle-image velocimetry data

Physics-Informed Neural Networks Enhanced Particle Tracking Velocimetry: An Example for Turbulent Jet Flow

Experimental dataset investigation of deep recurrent optical flow learning for particle image velocimetry: Flow past a circular cylinder

Adaptive-SpikeNet: Event-based Optical Flow Estimation using Spiking Neural Networks with Learnable Neuronal Dynamics

Deep Networks for Image-to-Image Prediction

On Using CFD and Experimental Data to Train an Artificial Neural Network to Reconstruct ECVT Images: Application for Fluidized Bed Reactors

Deep-learning-based image preprocessing for particle image velocimetry

226 Advancing Neurovascular Diagnostics for Abnormal Hemodynamic Conditions Through AI-Driven Physics-informed Neural Networks