Table 4.

Performance of the networks with different components on the Stanford dataset with Cartesian sampling.

Methods	NMSE (× 10−3) ↓		PSNR ↑		SSIM (× 10−2) ↑
	Mean±std	p-value	Mean±std	p-value	Mean±std	p-value
Without self-attention	11.95±0.01	7.4e-63	35.19±1.81	1.1e-59	92.04±0.02	2.0e-40
Channel-attention alone	11.52±0.01	9.7e-50	35.39±1.83	8.4e-48	92.15±0.03	8.1e-38
Frequency-attention alone	11.43±0.01	7.8e-37	35.43±1.87	5.3e-3	92.17±0.03	1.1e-17
ACNN-k-Space	11.11±0.01	-	35.55±1.86	-	92.25±0.03	-