Development of an image processing pipeline for the extraction of white matter from MRI

Abstract

newline On 100 test images, the acuity of the output images of the two-way diffusion, Adaptive Bilateral Filter (ABF), AKTV and LASKR, measured with the help of Maximum Local Variation (MLV) sharpness metric are 0.0634 ± 0.0041, 0.1218 ± 0.0213, 0.1263 ± 0.0368 and 0.1685 ± 0.0163. Standard Deviation (SD) of noise measured from the output images are 0.0676 ± 0.0158, 1.8654 ± 0.6203, 0.6285 ± 0.2140 and 0.4939 ± 0.2951. On100 low-field MRI slices used for performance assessment, the FCET exhibited Visual Information Fidelity (VIF) scores of 1.3340±0.2764 which is higher than those exhibited by ten state-of-the-art image contrast enhancement algorithms. On 100 MRI slices the Dice Similarity Index (DSI) shown by the Enhanced Fuzzy Segmentation Framework (EFSF), Modified FCM based on Double Estimation (MFCMDE), Spatial FCM (SFCM), Weighted SFCM (WSFCM), and PMSFCA in order are 0.7747 ± 0.0357, 0.38435 ± 0.246, 0.6212 ± 0.3655, 0.4207 ± 0.2784, 0.8300 ± 0.0158