. Author manuscript; available in PMC: 2022 Aug 1.

Published in final edited form as: Med Image Anal. 2021 May 16;72:102098. doi: 10.1016/j.media.2021.102098

Table 8.

Performance of our method with different numbers of the input slices based on the Stanford dataset.

Number of input slices	NMSE (× 10⁻³) ↓		PSNR ↑		SSIM (× 10⁻²) ↑		Train times	Test time (each case)
Number of input slices	Mean±std	p-value	Mean±std	p-value	Mean±std	p-value	Train times	Test time (each case)
	11.55±0.01	-	35.45±1.81	-	92.15±0.03	-	~44 hours	~49 ms
	11.11±0.01	7.4e-63	35.55±1.86	2.1e-53	92.25±0.03	6.3e-30	~50 hours	~50 ms
	10.94±0.00	1.3e-64	35.57±1.84	5.1e-61	92.30±0.03	2.2e-42	~57 hours	~58 ms
	10.76±0.00	2.0e-66	35.61±1.84	1.1e-61	92.35±0.03	8.8e-45	~65 hours	~62 ms
	10.65±0.00	1.1e-65	35.64±1.84	1.2e-61	92.40±0.03	2.4e-45	~74 hours	~65 ms
11	10.64±0.00	2.9e-66	35.64±1.84	1.1e-61	92.45±0.03	9.7e-46	~84 hours	~70 ms