- SOLIDWORKS FLOW SIMULATION REFINEMENT FIELD GRADIENT VERIFICATION
- SOLIDWORKS FLOW SIMULATION REFINEMENT FIELD GRADIENT PC
SOLIDWORKS FLOW SIMULATION REFINEMENT FIELD GRADIENT VERIFICATION
However, standalone CFD can also be limited due to its high dependence on patient-specific input conditions (boundary conditions), and its need for appropriate verification and validation. Furthermore, computational model conditions can be manipulated to match physiological or surgical variations of interest. Because CFD is computationally based, there is almost no limit to its flow resolution, given the appropriate computational resources. CFD is a method that utilizes the governing equations of fluid flow to calculate a velocity field from input geometric data and flow conditions. Another widely available cerebral blood flow analysis method, which may help address some of these limitations, is computational fluid dynamics (CFD) 10, 11, 12. For example, flow resolution limits can result from patient scan time restrictions and from imperfections that result from manipulation of a magnetic field (image noise, intravoxel dephasing, eddy currents, magnetic field non-linearity, etc.). 4D-flow can be used to non-invasively visualize and quantify blood flow in complex cerebrovascular systems however, flow analysis with MRI has some limitations when used as a stand-alone analysis method. One method that has emerged as a powerful, in-vivo technique is the direct measurement of blood velocities with four-dimensional flow (4D-flow) magnetic resonance imaging (MRI), which is a three-dimensional, time-resolved form of phase contrast (PC) MRI 8, 9. Such metrics can be derived or calculated with a variety of medical imaging and computational methods. Hemodynamic metrics can be of great value in cerebrovascular disease diagnosis and treatment planning 1, 2, 3, 4, 5, 6, 7.
SOLIDWORKS FLOW SIMULATION REFINEMENT FIELD GRADIENT PC
Such image enhancement can improve experimental and clinical qualitative and quantitative cerebrovascular PC MRI analysis. Through testing on simulated images, phantom data, and cerebrovascular 4D flow data from 20 patients, the trained network successfully de-noised flow images, decreased velocity error, and enhanced near-vessel-wall velocity quantification and visualization. In this preliminary study, the potential utility of such a method was demonstrated by using high resolution patient-specific CFD data to train a convolutional neural network, and then using the trained network to enhance MRI-derived velocity fields in cerebral blood vessel data sets.
Augmentation of cerebral 4D flow MRI data with CFD-informed neural networks may provide a method to produce highly accurate physiological flow fields. One method that excels in creating low-error, physics-based, velocity fields is computational fluid dynamics (CFD). However, limitations in both quantitative and qualitative analyses can result from errors inherent to PC MRI. Blood flow metrics obtained with four-dimensional (4D) flow phase contrast (PC) magnetic resonance imaging (MRI) can be of great value in clinical and experimental cerebrovascular analysis.