Level Set Segmentation of Medical Images Based on Local Region Statistics and Maximum a Posteriori Probability

Table 1

Comparison of the segmentation results obtained by the four methods over 20 images from the BrainWeb (mean ± SD).

Tissue	Chen’s method	Li’s method	Chen’s improved method	Proposed method
WM	0.8613 ± 0.0254	0.8792 ± 0.0223	0.9011 ± 0.0147	0.9289 ± 0.0152
GM	0.7583 ± 0.0317	0.7723 ± 0.0298	0.7953 ± 0.0241	0.8128 ± 0.0248
CSF	0.7349 ± 0.0414	0.7543 ± 0.0405	0.8311 ± 0.0319	0.8532 ± 0.0296