. Author manuscript; available in PMC: 2009 Oct 5.

Published in final edited form as: IEEE Trans Med Imaging. 2008 Oct;27(10):1389–1403. doi: 10.1109/TMI.2008.920609

TABLE II.

Quality Measures: SSIM, QILV, and MSE for the Images in the Experiments. Best Value of Each Column is Highlighted. LMMSE-Based Schemes Show Better Results in Terms of Noise Removal and Edge Preservation

	σ_n =5			σ_n = 10			σ_n = 20

	SSIM	QILV	MSE	SSIM	QILV	MSE	SSIM	QILV	MSE
Noisy	0.9235	0.9977	24.8971	0.7904	0.9890	100.2940	0.5722	0.9251	395.0881

CA	0.8544	0.6394	190.9824	0.8491	0.6430	192.1922	0.8236	0.6543	205.6214
EM	0.5726	0.0140	2839.1	0.8685	0.6513	144.1177	0.8373	0.6342	168.0615
Koay	0.8797	0.6590	137.6257	0.8854	0.6856	133.9855	0.8280	0.5527	179.0641
ML	0.3770	0.2265	2120.9	0.8681	0.6516	144.2500	0.8370	0.6354	168.1712
Gaussian	0.8904	0.6479	128.5180	0.8789	0.6073	139.1106	0.8392	0.5145	190.3817
Wiener	0.9664	0.9967	18.1872	0.9092	0.9839	57.9197	0.8146	0.9076	161.8120

LMMSE	0.9681	0.9980	17.7973	0.9168	0.9921	53.9731	0.8346	0.9613	130.5361
RLMMSE (8)	0.9713	0.9981	17.4090	0.9270	0.9917	51.8197	0.8597	0.9502	122.5699
RLMMSE (50)	0.9714	0.9982	17.4562	0.9298	0.9915	51.8487	0.8540	0.9429	129.5132