Skip to main content
. 2023 Feb 20;13(4):794. doi: 10.3390/diagnostics13040794

Figure 1.

Figure 1

Maximum-intensity projection of (a) conventional positron emission tomography (PET) reconstruction and (b) deep learning (DL) PET restoration of a woman in her 60 s with invasive ductal carcinoma in the right breast (arrows): invasive diameter, 15 mm; nuclear grade, 2; lymphatic invasion, negative; vascular invasion, negative; estrogen receptor, positive; progesterone receptor, positive; HER2 receptor, positive; Ki-67 staining, 25.5%; and nodal stage, 0. Axial view of a tumor cross-section for the same woman using (c) cPET reconstruction and (d) DL-PET restoration. Images were scored on a five-point scale from 1 = extremely poor to 5 = excellent. The scores of reader 1 for “depiction of the primary lesion”, “noise”, “clarity of mammary gland”, and “overall image quality” were 4, 5, 5, and 5 for cPET and 5, 5, 5, and 5 for DL-PET, respectively; the scores of reader 2 were 3, 4, 4, and 4 for cPET and 5, 5, 5, and 5 for DL-PET, respectively.