Skip to main content

Table 2 Overview over all u-nets trained in this study

From: Analysis of a deep learning-based method for generation of SPECT projections based on a large Monte Carlo simulated dataset

Abbreviation Poisson noise realization Number of input/output projections Cyclical expansion Rotation shift Size of training/validation dataset
U1 A 60 64 9000/500
U2 B 60 64 9000/500
U3 A 60 64 400/40
U4 A 30 32 9000/500
U5 A 30 32 9000/500
U6 A 30 32 9000/500
U7 A 120 120 9000/500