A Clustering Approach to Improve IntraVoxel Incoherent Motion Maps from DW-MRI Using Conditional Auto-Regressive Bayesian Model

The Intra-Voxel Incoherent Motion (IVIM) model allows to estimate water diffusion and perfusion-related coefficients in biological tissues using weighted MR images. Among the available approaches fit IVIM bi-exponential decay, a segmented Bayesian algorithm with Conditional Auto-Regressive (CAR) prior spatial regularization has been recently proposed produce more reliable coefficient estimation. However, CAR can generate inaccurate estimation, especially at interfaces between different tissues. To overcome this problem, was coupled work k-means clustering approach, separate exclude voxels from other regions specification. approach compared original method without state-of-the-art CAR. were tested on simulated images by calculating estimation error of variation (CV). Furthermore, applied some illustrative real oncologic patients. On images, innovation reduced average 47%, 21% 58% for D, f D*, respectively, 48% 34% D f, CAR, while it achieved same D*. also able consistently reduce CV each coefficient. novel did not alter maps obtained method, advantage reducing their typical blotchy appearance boundaries. represents valuable improvement over provides is less sensitive bias inconsistency tissue/tissue tissue/background interfaces.