Figure 2 | Scientific Reports

Figure 2

From: A trial deep learning-based model for four-class histologic classification of colonic tumor from narrow band imaging

Figure 2

The patch-level histological mapping predicted by the trained model, alongside the annotation mask. An example picture of SMd and the annotation mask without blackout or halation can be seen in (a) and (i), respectively. White (b), green (c), yellow (d), magenta (e), red (f), dark gray (g), and cyan (h) open squares indicate the patches classified as BG, LGD, HGD, SMs, SMd, BG-oof, and L-oof, respectively, with their corresponding union masks visible in (j–p), respectively. BG, background; HGD, high grade dysplasia; LGD, low grade dysplasia; SMs, superficially invasive submucosal carcinoma; SMd, deeply invasive submucosal carcinoma; BG-oof, out-of-focus background; L-oof, out-of-focus lesion.

Back to article page