Method of Magnetic Resonance Imaging Reconstruction Based on Lp-Norm Joint Total Variation

Jizhong Duan; Xiaoxun He; Chang Liu; Minghong Xie

doi:10.3788/LOP202158.2411001

Journals >Laser & Optoelectronics Progress >Volume 58 >Issue 24 >Page 2411001 > Article

Laser & Optoelectronics Progress
Vol. 58, Issue 24, 2411001 (2021)

Method of Magnetic Resonance Imaging Reconstruction Based on L_p-Norm Joint Total Variation

Jizhong Duan^*, Xiaoxun He, Chang Liu, and Minghong Xie

Author Affiliations

Faculty of Information Engineering and Automation, Kunming University of Science and Technology, Kunming, Yunnan 650504, China

show less

DOI: 10.3788/LOP202158.2411001 Cite this Article Set citation alerts

Jizhong Duan, Xiaoxun He, Chang Liu, Minghong Xie. Method of Magnetic Resonance Imaging Reconstruction Based on L_p-Norm Joint Total Variation[J]. Laser & Optoelectronics Progress, 2021, 58(24): 2411001 Copy Citation Text

show less

Reconstructed images and error graphs of different algorithms under 6-fold acceleration on GE_human_brain dataset. (a) Original images; (b) P-LORAKS algorithm; (c) JTV-PLORAKS algorithm; (d) proposed algorithm

Fig. 1. Reconstructed images and error graphs of different algorithms under 6-fold acceleration on GE_human_brain dataset. (a) Original images; (b) P-LORAKS algorithm; (c) JTV-PLORAKS algorithm; (d) proposed algorithm

Download full size

Reconstructed images and error graphs of different algorithms under 6-fold acceleration on P1_imk_100 dataset. (a) Original images; (b) P-LORAKS algorithm; (c) JTV-PLORAKS algorithm; (d) proposed algorithm

Fig. 2. Reconstructed images and error graphs of different algorithms under 6-fold acceleration on P1_imk_100 dataset. (a) Original images; (b) P-LORAKS algorithm; (c) JTV-PLORAKS algorithm; (d) proposed algorithm

Download full size

Fig. 3. Reconstructed images and error graphs of different algorithms under 8-fold acceleration on GE_human_brain dataset. (a) Original images; (b) P-LORAKS algorithm; (c) JTV-PLORAKS algorithm; (d) proposed algorithm

Download full size

Fig. 4. Reconstructed images and error graphs of different algorithms under 8-fold acceleration on P1_imk_100 dataset. (a) Original images; (b) P-LORAKS algorithm; (c) JTV-PLORAKS algorithm; (d) proposed algorithm

Download full size

Algorithm	Index	R'=3	R'=4	R'=5	R'=6	R'=7	R'=8
P-LORAKS	SNR /dB	27.5100	25.7100	23.8000	21.8500	20.7700	19.3700
	NRMSE	0.0059	0.0073	0.0091	0.0114	0.0129	0.0152
	HFEN	0.0599	0.0760	0.0972	0.1262	0.1463	0.1756
	SSIM	0.9620	0.9438	0.9200	0.8838	0.8591	0.8280
JTV_PLORAKS	SNR /dB	28.4400	26.6500	25.2200	23.6700	22.3800	21.2400
	NRMSE	0.0053	0.0065	0.0077	0.0092	0.0107	0.0122
	HFEN	0.0565	0.0729	0.0925	0.1194	0.1444	0.1681
	SSIM	0.9688	0.9581	0.9486	0.9339	0.9228	0.9083
Proposed algorithm	SNR /dB	28.9000	26.9800	25.5300	24.2000	23.4900	22.3600
	NRMSE	0.0050	0.0063	0.0075	0.0087	0.0094	0.0108
	HFEN	0.0575	0.0714	0.0806	0.1070	0.1192	0.1415
	SSIM	0.9718	0.9604	0.9493	0.9393	0.9323	0.9208

Table 1. Reconstruction effects of three algorithms in GE_human_brain dataset

Algorithms	Index	R'=3	R'=4	R'=5	R'=6	R'=7	R'=8
P-LORAKS	SNR /dB	14.0600	11.6300	10.3200	9.3200	8.5200	8.2300
	NRMSE	0.0249	0.0330	0.0384	0.0435	0.0469	0.0488
	HFEN	0.3244	0.4456	0.5177	0.5883	0.6415	0.6872
	SSIM	0.6846	0.5765	0.4981	0.4372	0.4042	0.3822
JTV_PLORAKS	SNR /dB	14.9000	12.7100	12.1800	11.0900	10.5800	9.9700
	NRMSE	0.0226	0.0291	0.0310	0.0351	0.0372	0.0400
	HFEN	0.3333	0.4572	0.4984	0.5800	0.6519	0.6885
	SSIM	0.7036	0.6051	0.5711	0.5171	0.4916	0.4669
Proposed algorithm	SNR /dB	15.5700	13.4300	12.9600	12.4300	12.1100	11.6100
	NRMSE	0.0210	0.0268	0.0283	0.0301	0.0312	0.0331
	HFEN	0.2762	0.3776	0.4374	0.4764	0.5203	0.5507
	SSIM	0.7814	0.7253	0.6681	0.6551	0.6327	0.6181

Table 2. Reconstruction effects of three algorithms in P1_imk_100 dataset

Download Citation

Set citation alerts for the article

Tools

Set citation alerts for the article

Save the article for my favorites

Paper Information