Figure - PMC

Skip to main content

An official website of the United States government

Here's how you know

Here's how you know

Official websites use .gov
A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS
A lock ( ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

View full-text article in PMC

. 2022 Nov 2;12:18466. doi: 10.1038/s41598-022-22731-x

Search in PMC
Search in PubMed
View in NLM Catalog
Add to search

© The Author(s) 2022

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

PMC Copyright notice

Overview of EBVNet. EBVNet was trained on the TCGA dataset, and fine-tuned on the ISH dataset. EBVNet before and after fine-tuning were externally validated on the HGH dataset. Each H&E histology slide was preprocessed by discarding non tissue-containing background using Otsu’s thresholding and generating non-overlapping patches. All the patches were stain normalized (not shown in the figure). First, the tumor classification model (blue box) receives all the tissue-containing patches and returns the probability of each patch being a tumor patch. The patches with probabilities of higher than 0.5 are assigned as tumor patches, otherwise, they are assigned as normal patches. The tumor-predicted patches are then fed into the second classifier (EBV prediction model, purple box), with the resulting probability of being an EBV positive tumor patch. The tumor-predicted patches with probabilities higher than 0.1 are assigned as EBV positive tumor patches, otherwise, they are assigned as EBV negative patches. Based on the patch-wise classification result, the EBV probability score (EPS), which is the ratio of the number of EBV-positive tumor-predicted patches to the number of tumor-predicted patches, is calculated for each slide. The slide with calculated EPS higher than 0.2 is assigned as an EBV positive slide, otherwise, it is assigned as an EBV negative slide.