Skip to main content
. 2022 Jan 11;11:802205. doi: 10.3389/fonc.2021.802205

Figure 1.

Figure 1

The design flow chart of this paper. (A) Acquisition of ROI images for Borrmann type IV GC and PGL. (B) The ROI features extraction by Densenet121. (C) Transfer learning features selection process and model building, statistical analysis of clinical characteristics, and CT subjective findings. (D) Performance evaluation of the transfer learning radiomics nomogram. GC, gastric cancer; PGL, primary gastric lymphoma; ROI, region of interest.