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

Figure 3

Segmentation results on a T1-weighted sagittal slice (number 100) from SBD with 7% noise and 20% grayscale nonuniformity. (a) Original image. (b) Ground truth. (c) GKFCM1 results. (d) GKFCM2 results. (e) FLICM results. (f) KWFLICM results. (g) MICO results. (h) RSCFCM results. (i) ARKFCM₁ results. (j) ARKFCM₂ results. (k) results.

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

(j)

(k)