AB012. Automated histologic subtyping of thymic epithelial tumors with deep learning
Abstract

AB012. Automated histologic subtyping of thymic epithelial tumors with deep learning

James M. Dolezal1, Wenji Guo1, Christine Bestvina1, Everett Vokes1, Jessica Donington2, Aliya Husain3, Marina Chiara Garassino1

1Section of Hematology/Oncology, Department of Medicine, University of Chicago Medicine, Chicago, IL, USA; 2Department of Pathology, University of Chicago Medicine, Chicago, IL, USA; 3Section of Thoracic Surgery, University of Chicago Medicine, Chicago, IL, USA

Correspondence to: James M. Dolezal. Section of Hematology/Oncology, Department of Medicine, University of Chicago Medicine, 5841 S Maryland Ave., Chicago, IL 60637, USA. Email: james.dolezal@uchospitals.edu.

Background: Rare tumors are diagnostic challenges for pathologists. Thymic epithelial tumors (TETs) are heterogenous and their treatment strategies vary according to histological subgroup. Previous work has shown that a second pathological opinion may result in a change in diagnosis for more than half of cases, with a potential treatment shift in 44%. The aim of this study is to assess the feasibility of using artificial intelligence and deep learning to classify TETs, which could be used to help improve pathologist diagnostic consistency for these challenging tumors.

Methods: Digital diagnostic hematoxylin and eosin (H&E) stained slides of tumors for 103 patients with thymoma type A, AB, B1, B2, and B3 were downloaded from The Cancer Genome Atlas (TCGA). An Xception-based deep convolutional neural network model was trained on slide images at 10× magnification to predict histologic subtype as an ordinal variable in three-fold cross-validation. Hyperparameters were taken from previously published experiments, and no additional hyperparameter tuning was performed to reduce the risk of overfitting. Validation predictions from each cross-fold were aggregated and compared between groups using analysis of variance (ANOVA) and one-sided t-tests with Bonferroni correction for multiple comparisons. Model activations at the post-convolutional layer for validation images in the first cross-fold were visualized with uniform manifold approximation and projection (UMAP) dimensionality reduction to better understand the spatial relationship between learned image features.

Results: Deep learning predictions among the TET subtypes were significantly different by ANOVA (P<0.0001) and correlated with the ordinal labels (R-squared =0.39). Thymoma A and AB subtypes were distinguished from both B1 and B2/B3 (P=0.023 and <0.001, respectively), and B1 tumors were distinguished from B2/B3 (P=0.011). Analysis of post-convolutional layer activations revealed an axis of transition through the ordinal variables, providing evidence that the deep learning model learned image features on a morphologic spectrum.

Conclusions: This is the first example in TETs that deep learning can discriminate between TET histologic subtypes using digital H&E slides. We aim to further validate the algorithm with a multi-institution dataset from centers of expertise to improve the ability to distinguish thymoma subtypes.

Keywords: Deep learning; computational pathology; thymic epithelial tumor (TET)

doi: 10.21037/med-22-ab012

Acknowledgments

Funding: The project described was supported by Grant Number 5 T32 CA 9566-34 from NIH. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH.

Footnote

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://med.amegroups.com/article/view/10.21037/med-22-ab012/coif). JMD reports that the project described was supported by Grant Number 5 T32 CA 9566-34 from NIH. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH. CB reports grants or contracts from AstraZeneca, BMS; consulting fees from AstraZeneca, BMS, CVS, Genentech, Jazz, JNJ, Novartis, Pfizer, Regeneron, Sanofi, Seattle Genetics, Takeda; Payment or honoraria for speaker from Merck. EV reports consulting fees from AstraZeneca, Beigene, BioNTech, BMS, Eli Lilly, EMD Serono, GeneCentric, Genentech/Roche, GlaxoSmithKline, Novartis; payment or honoraria from AstraZeneca, Beigene, BioNTech, BMS, Eli Lilly, EMD Serono, GeneCentric, Genentech/Roche, GlaxoSmithKline, Novartis; leadership or fiduciary role in ASCO; stock or stock options of McKesson, Coordination Pharmaceuticals. JD reports honoraria for speaking and advisory board from AstraZeneca, BMS, Roche/Genentech, and Merk. MCG reports consulting fees from AstraZeneca, MSD International GmbH, BMS, Boehringer Ingelheim Italia S.p.A, Celgene, Eli Lilly, Ignyta, Incyte, Inivata, MedImmune, Novartis, Pfizer, Roche, Takeda, Seattle Genetics, Mirati, Daiichi Sankyo, Regeneron, Merck, Ose Immuno Therapeutics, Blueprint, Jansenn, and Sanofi; payment or honoraria from Eli Lilly, MSD, Pfizer (MISP), AstraZeneca, MSD International GmbH, BMS, Boehringer Ingelheim Italia S.p.A, Celgene, Ignyta, Incyte, MedImmune, Novartis, Pfizer, Roche, Takeda, Tiziana, Foundation Medicine, Glaxo Smith Kline GSK, Spectrum pharmaceuticals, unrelated to the present manuscript. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

doi: 10.21037/med-22-ab012
Cite this abstract as: Dolezal JM, Guo W, Bestvina C, Vokes E, Donington J, Husain A, Garassino MC. Automated histologic subtyping of thymic epithelial tumors with deep learning. Mediastinum 2022;6:AB012.

Article Options

Download Citation

Share