Unsupervised Approach Data Analysis Based on Fuzzy Possibilistic Clustering: Application to Medical Image MRI

Table 7

Results comparison for clustering index (Se and Sp) for GM tissue by FCM, PCM, PFCM, and HPFCM, with different noises (0%, 1%, 3%, 5%, 7%, and 9%), using Brainweb data sets.

Noise	FCM		PCM		PFCM		HPFCM
	Se	Sp	Se	Sp	Se	Sp	Se	Sp
0%	85.19	89.15	85.93	87.71	96.33	96.61	96.69	97.73
1%	83.73	88.28	84.75	86.63	96.18	95.42	96.11	96.45
3%	78.92	85.17	83.28	85.32	94.81	95.05	95.82	96.13
5%	73.45	83.09	82.09	84.14	92.77	94.58	95.13	95.62
7%	70.20	81.72	80.69	83.31	91.39	93.98	94.05	94.81
9%	65.73	80.75	78.63	82.19	90.26	92.28	92.97	93.72