Fig. 5From: Sequential deep learning image enhancement models improve diagnostic confidence, lesion detectability, and image reconstruction time in PETQuantitative performance of five reconstructions of DLE/DLT combinations. Tested by drawing a 30 mm spherical VOI in both the lung and liver and measuring standard deviation between voxels and SUVmean for 20 patients scanned on a GE HealthCare Discovery 710 scannerBack to article page