Be first to read the latest tech news, Industry Leader's Insights, and CIO interviews of medium and large enterprises exclusively from Medical Tech Outlook
THANK YOU FOR SUBSCRIBING
By
MedTech Outlook | Monday, September 25, 2023
Machine learning can make 4D flow MRI more efficient, accurate, and accessible.
FREMONT, CA: Magnetic Resonance Imaging (MRI) has revolutionized medical diagnostics by providing non-invasive, high-resolution images of internal structures. One particularly promising MRI technique, 4D flow MRI, is becoming increasingly useful in diagnosing and understanding cardiovascular diseases. It enables the quantification of net flow, peak velocities, and visualization of blood flow direction, presenting vital information for conditions such as stenoses, aortic coarctation, aortic and mitral valve regurgitation, aortic aneurysms, hypertrophy cardiomyopathy, and congenital heart diseases.
Stay ahead of the industry with exclusive feature stories on the top companies, expert insights and the latest news delivered straight to your inbox. Subscribe today.
The ability of 4D flow MRI to quantify net flow and peak velocities has proven invaluable in cardiovascular disease assessment. This technology aids in grading stenoses, assessing aortic coarctation severity, and evaluating aortic and mitral valve regurgitation. Additionally, 4D flow MRI facilitates the visualization of blood flow direction, providing crucial insights for conditions like aortic aneurysms, aortic dissections, and coarctations, hypertrophy cardiomyopathy, and congenital heart diseases involving univentricular hearts or transposition of the great arteries. Notably, it allows direct quantification of regurgitant flow, significantly improving over traditional indirect methods used in mitral valve insufficiency assessment.
Beyond diagnostic purposes, 4D flow MRI has become instrumental in analyzing complex flow patterns pre and post-surgery, such as in the Fontan procedure. Moreover, additional biomarkers like kinetic energy (KE), turbulent kinetic energy (TKE), viscous energy (VE) loss, wall shear stress (WSS), and pulse wave velocity (PWV) show promise in distinguishing patients with cardiovascular diseases from healthy subjects.
Despite its immense potential, 4D flow MRI does face several limitations that hinder its widespread clinical adoption. The most notable challenge is its lengthy acquisition time, typically 10 minutes, four times longer than conventional cine MRI scans. The extended scan time increases costs and leads to patient discomfort and susceptibility to motion artifacts.
Another drawback lies in limited spatiotemporal resolutions caused by the signal-to-noise ratio (SNR) and scan time constraints. This limitation results in velocity measurement inaccuracies, especially in small vessels, low-flow venous vessels, and near vessel walls due to partial volume effects. Consequently, stenosis grading and wall shear stress estimation may suffer from inaccuracies.
Inherent inaccuracies in the MRI measurement process, such as residual phase errors induced by various factors, can lead to velocity estimation errors. Correcting these inaccuracies requires time-consuming post-processing using dedicated software, where the manual placement of 2D planes and contours is common for evaluating net flow and peak velocities. Similarly, parameters like KE, VE, TKE, and WSS necessitate meticulous 3D vessel lumen delineation.
In response to these challenges, integrating machine learning (ML) techniques has emerged as a promising solution to enhance 4D flow MRI's capabilities. Scan acceleration methods, like compressed sensing (CS), have been extended using ML reconstructions, drastically reducing image reconstruction time to a few seconds. ML super-resolution techniques incorporated high-resolution computational fluid dynamics simulations and 4D flow MRI to achieve more realistic velocity results.
More in News
I agree We use cookies on this website to enhance your user experience. By clicking any link on this page you are giving your consent for us to set cookies. More info

However, if you would like to share the information in this article, you may use the link below:
www.medicaltechoutlookeurope.com/news/mri-advances-in-diagnosis-and-treatment-nwid-2993.html
