MR Image Reconstruction Based on Iterative Split Bregman Algorithm and Nonlocal Total Variation

Table 1

Comparison of the average values of quality evaluation parameters (Gaussian noise case).

	Image	TVCMRI	RecPF	CSA	FCSA	SB-TV	Proposed
SNR (dB)	Brain	14.19	14.75	15.20	15.78	16.31	17.49
	Chest	16.99	17.38	17.13	17.63	18.03	18.89
	Artery	18.70	19.98	22.15	23.61	24.17	25.75
	Cardiac	16.97	17.65	17.95	18.77	18.98	19.47
RE (%)	Brain	16.96	16.54	07.97	07.37	06.90	05.98
	Chest	11.58	11.21	10.84	10.33	08.89	07.96
	Artery	12.61	11.84	05.77	04.90	04.57	04.10
	Cardiac	14.03	13.39	11.10	09.78	08.29	07.72
SSIM	Brain	0.9567	0.9742	0.9906	0.9935	0.9940	0.9947
	Chest	0.9829	0.9839	0.9848	0.9861	0.9887	0.9914
	Artery	0.9825	0.9835	0.9848	0.9919	0.9924	0.9940
	Cardiac	0.9614	0.9642	0.9748	0.9799	0.9835	0.9889
FSIM	Brain	0.8712	0.8773	0.9489	0.9654	0.9719	0.9765
	Chest	0.8990	0.9012	0.9063	0.9148	0.9237	0.9374
	Artery	0.9010	0.9168	0.9516	0.9668	0.9705	0.9742
	Cardiac	0.9313	0.9366	0.9454	0.9551	0.9587	0.9629
CPU time (s)	Brain	2.14	2.05	1.96	1.88	1.78	1.62
	Chest	2.21	2.16	1.95	1.86	1.73	1.62
	Artery	2.26	2.15	1.86	1.83	1.72	1.64
	Cardiac	2.13	2.10	1.84	1.82	1.80	1.67