End-to-end learning for detecting MYC translocations

Stephan Dooper, Geert Litjens

Keywords: MYC, DLBCL, End-to-end learning, whole slide image classification

TL;DR: We use an end-to-end whole slide classification algorithm to classify MYC translocations in H&E slides of DLBCL

Abstract: Recent developments have improved whole-slide image classification to the point where the entire slide can be analyzed using only weak labels, whilst retaining both local and global context. In this paper, we use an end-to-end whole-slide image classification approach using weak labels to classify MYC translocations in slides of diffuse large B-cell lymphoma. Our model is able to achieve an AUC of 0.8012, which indicates the possibility of learning relevant features for MYC translocations.

Registration: I acknowledge that acceptance of this work at MIDL requires at least one of the authors to register and present the work during the conference.

Authorship: I confirm that I am the author of this work and that it has not been submitted to another publication before.

Paper Type: novel methodological ideas without extensive validation

Primary Subject Area: Application: Histopathology

Secondary Subject Area: Learning with Noisy Labels and Limited Data

Confidentiality And Author Instructions: I read the call for papers and author instructions. I acknowledge that exceeding the page limit and/or altering the latex template can result in desk rejection.