Table 2.

Region-based analyses: the mean, standard deviation (SD), and the 95% confidence interval (CI) of the regional standardized uptake value ratio (SUVR) differences between various image types and the ground-truth images; Cohen's d effect sizes; area under the curves (AUC); 95% CI of the AUC difference between the DL-based methods and the ground truth.

Image Type	Mean±SD, 95% CI	Cohen’s d	AUC	95% CI of difference AUC
Ground-truth		2.25	0.954	N/A
Method A	0.023±0.172) (−0.321, 0.367)	2.38	0.982	(−0.02, 0.08)
Method B	−0.010±0.093 (−0.196, 0.176)	2.64	0.995	(−0.02, 0.10)
Method C	−0.013±0.094 (−0.201, 0.175)	2.33	0.987	(−0.02, 0.09)
Method D	−0.016±0.098 (−0.212, 0.180)	2.59	0.991	(−0.02, 0.09)
Low–count data	−0.004±0.130 (−0.264, 0.256)	1.79	0.945	N/A

Network training methods: A-direct application of pre-trained network; B-transfer learning starting with pre-trained network; C-training new network from scratch; D-training new network with combined datasets