Four machine learning inversion algorithms with different material spatial property assumptions (trained on simulated data with homogeneous, piecewise constant, smooth, or piecewise smoothly varying material properties) were evaluated in a brain simulating phantom with stiff inclusions, a test-retest repeatability dataset, and an Alzheimer’s disease dataset. The piecewise smooth inversion produced the highest contrast to noise ratio and allowed the best visualization of inclusions in the phantom study. All four inversions produced stiffness estimates that were repeatable and sensitive to stiffness changes in Alzheimer’s disease.
This abstract and the presentation materials are available to members only; a login is required.