Table 5.

Quantitative evaluation (MEAN ± SD) for 400 test images from the AR dataset. A detailed explanation of all used metrics can be found in Table 2. The regions corresponding to the labels “bk” and “#1”–“#5” can be found in Fig. 5.

	Algorithms	Tissue-wise metrics						tSD	tMRE (top) tPSNR (bottom)
		bk	#1	#2	#3	#4	#5
$\left|\text{Relative error}\right|$	Unet-MSE	0.148 ± 0.064	0.040 ± 0.022	0.043 ± 0.025	0.092 ± 0.051	0.108 ± 0.059	0.101 ± 0.047	0.038	0.088 ± 0.022
	Unet-MSRE	0.085 ± 0.025	0.075 ± 0.025	0.071 ± 0.024	0.100 ± 0.044	0.110 ± 0.054	0.139 ± 0.050	0.023	0.097 ± 0.019
	Ours	0.035 ± 0.018	0.068 ± 0.031	0.066 ± 0.030	0.082 ± 0.042	0.072 ± 0.040	0.073 ± 0.042	0.015	0.066 ± 0.018
PSNR (dB)	Unet-MSE	12.214 ± 2.704	26.052 ± 3.277	25.764 ± 3.420	20.187 ± 3.597	18.748 ± 2.844	18.150 ± 2.844	4.749	20.186 ± 1.555
	Unet-MSRE	19.345 ± 2.242	16.678 ± 1.481	17.533 ± 1.501	17.816 ± 2.267	18.143 ± 2.952	13.991 ± 1.654	1.660	17.251 ± 1.137
	Ours	23.868 ± 2.445	19.518 ± 2.297	20.045 ± 2.241	20.243 ± 3.037	21.864 ± 3.402	21.133 ± 3.438	1.449	21.112 ± 1.592