Multispectral texture characterization: application to computer aided diagnosis on prostatic tissue images

EURASIP Journal on Advances in Signal Processing

Table 2 Confusion matrices obtained by using (a) GLCM (32 gray levels), (b) GLCM (64 gray level), (c) SSGLDM (S = 16 bands, 32 gray level), (d) SSGLDM (S = 16 bands, 64 gray level), (e) SSGLDM (S = 3 bands, 32 gray levels), (f) SSGLDM (S = 3 bands, 64 gray levels)

Classified as:	BPH	PCa	PIN	Stroma	Error (%)
(a)
BPH	135	1	24	0	15.62
Pca	0	175	0	1	0.57
PIN	2	0	141	1	2.08
Stroma	0	0	3	141	2.08
Overall					5.13
(b)
BPH	138	0	22	0	13.75
Pca	0	176	0	0	0
PIN	3	0	141	0	2.08
Stroma	0	0	3	141	2.08
Overall					4.49
(c)
BPH	149	0	11	0	6.88
Pca	0	173	0	3	1.70
PIN	0	0	136	8	5.56
Stroma	0	11	5	128	11.11
Overall					6.09
(d)
BPH	145	0	15	0	9.38
Pca	0	174	0	2	1.14
PIN	0	0	135	9	6.25
Stroma	5	4	7	128	11.11
Overall					6.73
(e)
BPH	147	0	13	0	8.13
Pca	0	175	0	1	0.57
PIN	0	0	144	0	0
Stroma	0	0	0	144	0
Overall					2.24
(f)
BPH	147	0	13	0	7.5
Pca	0	175	0	1	0.57
PIN	0	0	143	1	0.69
Stroma	0	0	1	143	0.69
Overall					2.56