Please use this identifier to cite or link to this item: doi:10.22028/D291-41810
Title: Glioma subtype classification from histopathological images using in-domain and out-of-domain transfer learning: An experimental study
Author(s): Despotovic, Vladimir
Kim, Sang-Yoon
Hau, Ann-Christin
Kakoichankava, Aliaksandra
Klamminger, Gilbert Georg
Borgmann, Felix Bruno Kleine
Frauenknecht, Katrin B.M.
Mittelbronn, Michel
Nazarov, Petr V.
Language: English
Title: Heliyon
Volume: 10
Issue: 5
Publisher/Platform: Elsevier
Year of Publication: 2024
Free key words: Digital pathology
Whole slide images
Glioma
Deep learning
Transfer learning
DDC notations: 610 Medicine and health
Publikation type: Journal Article
Abstract: We provide in this paper a comprehensive comparison of various transfer learning strategies and deep learning architectures for computer-aided classification of adult-type diffuse gliomas. We evaluate the generalizability of out-of-domain ImageNet representations for a target domain of histopathological images, and study the impact of in-domain adaptation using self-supervised and multi-task learning approaches for pretraining the models using the medium-to-large scale datasets of histopathological images. A semi-supervised learning approach is furthermore proposed, where the fine-tuned models are utilized to predict the labels of unannotated regions of the whole slide images (WSI). The models are subsequently retrained using the ground-truth labels and weak labels determined in the previous step, providing superior performance in comparison to standard in-domain transfer learning with balanced accuracy of 96.91% and F1-score 97.07%, and minimizing the pathologist’s efforts for annotation. Finally, we provide a visualization tool working at WSI level which generates heatmaps that highlight tumor areas; thus, providing insights to pathologists concerning the most informative parts of the WSI.
DOI of the first publication: 10.1016/j.heliyon.2024.e27515
URL of the first publication: https://doi.org/10.1016/j.heliyon.2024.e27515
Link to this record: urn:nbn:de:bsz:291--ds-418101
hdl:20.500.11880/37402
http://dx.doi.org/10.22028/D291-41810
ISSN: 2405-8440
Date of registration: 27-Mar-2024
Faculty: M - Medizinische Fakultät
Department: M - Pathologie
Professorship: M - Prof. Dr. Rainer M. Bohle
Collections:SciDok - Der Wissenschaftsserver der Universität des Saarlandes

Files for this record:
File Description SizeFormat 
1-s2.0-S2405844024035461-main.pdf1,86 MBAdobe PDFView/Open


This item is licensed under a Creative Commons License Creative Commons