Energy enhanced tissue texture in spectral computed tomography for lesion classification

Visual Computing for Industry, Biomedicine, and Art

Table 3 RMSE (10e-4), PSNR, SSIM, FSIM Index for region of interest images in Fig. 7

Metrics	Methods	Reconstructed images (channel number)
Metrics	Methods	1th	2th	3th	4th	5th
RMSE	SART	6.5373	72,168	9.1134	9.8020	10.6390
	TV	6.3068	5.7223	5.9562	6.0101	5.9182
	LRTV	6.4144	5.5096	5.5894	5.4789	5.3151
	TDL	6.7663	5.4743	5.4770	5.2326	4.9444
	LRTP	4.7431	3.8528	3.8517	4.0398	4.2506
PSNR	SART	30.624	29.054	27.027	26.270	25.401
	TV	30.922	31.069	30.721	30.518	30.495
	LRTV	30.775	31.398	31.273	31.322	31.429
	TDL	30.311	31.453	31.449	31.721	32.057
	LRTP	33.397	34.505	34.507	33.969	33.370
SSIM	SART	0.9994	0.9993	0.9989	0.9987	0.9984
	TV	0.9992	0.9993	0.9993	0.9993	0.9993
	LRTV	0.9992	0.9993	0.9993	0.9994	0.9994
	TDL	0.9992	0.9994	0.9994	0.9995	0.9995
	LRTP	0.9995	0.9997	0.9997	0.9997	0.9997
FSIM	SART	0.9981	0.9975	0.9958	0.9947	0.9920
	TV	0.9950	0.9941	0.9942	0.9950	0.9952
	LRTV	0.9944	0.9943	0.9942	0.9948	0.9952
	TDL	0.9963	0.9970	0.9970	0.9973	0.9975
	LRTP	0.9988	0.9992	0.9992	0.9991	0.9989

TV Total variation, LRTV Low rand total variation, TDL Tensor dictionary learning, SART Simultaneous algebraic reconstruction technique, LRTP Low rank texture prior, RMSE Root of mean square error, PSNR Peak signal to noise ratio, SSIM Structure similarity index, FSIM Feature similarity index