Segmentation of Brain Tissues from Magnetic Resonance Images Using Adaptively Regularized Kernel-Based Fuzzy -Means Clustering

Table 2

JS and running time on the T1-weighted sagittal slice (number 100) from SBD with 7% noise and 20% grayscale nonuniformity.

Algorithm	GKFCM1	GKFCM2	FLICM	KWFLICM	MICO	ARKFCM1	ARKFCM2
WM	0.773	0.775	0.771	0.765	0.672	0.785	0.788	0.786
GM	0.796	0.806	0.794	0.791	0.669	0.815	0.816	0.816
CSF	0.834	0.852	0.835	0.824	0.869	0.871	0.872	0.875
Average	0.801	0.811	0.800	0.794	0.736	0.824	0.825	0.825
Time (s)	1.377	1.797	5.392	166.717	0.751	0.348	0.329	0.322