Improved Complex Convolutional Neural Network Based on SPIRiT and Dense Connection for Parallel MRI Reconstruction

Table 7

Comparisons of average PSNR and SSIM values with standard deviation of the MR images reconstructed by SPIRiT, DONet, DONet with SCB, and DONet with SCB and dense on the Coronal-PD dataset under the 2DPU pattern.

Reconstruction model	Threefold AF		Fivefold AF
	PSNR (dB)	SSIM	PSNR (dB)	PSNR (dB)
SPIRiT	39.398 ± 2.762	0.928 ± 0.026	36.832 ± 2.450	0.883 ± 0.034
DONet	39.786 ± 2.846	0.936 ± 0.025	37.597 ± 2.791	0.898 ± 0.037
DONet with SCB	40.650 ± 1.636	0.941 ± 0.012	38.674 ± 1.442	0.917 ± 0.014
DONet with SCB and dense	40.855 ± 1.722	0.942 ± 0.013	38.885 ± 1.557	0.919 ± 0.015

The best values are in bold.