부산대학교 도서관

Keywords: digital pathology; microsatellite instability; deep learning; generative adversarial network; generative model; colorectal cancer; artificial intelligence; machine learning Abstract Deep learning can detect microsatellite instability (MSI) from routine histology images in colorectal cancer (CRC). However, ethical and legal barriers impede sharing of images and genetic data, hampering development of new algorithms for detection of MSI and other biomarkers. We hypothesized that histology images synthesized by conditional generative adversarial networks (CGANs) retain information about genetic alterations. To test this, we developed a 'histology CGAN' which was trained on 256 patients (training cohort 1) and 1457 patients (training cohort 2). The CGAN synthesized 10000 synthetic MSI and non-MSI images which contained a range of tissue types and were deemed realistic by trained observers in a blinded study. Subsequently, we trained a deep learning detector of MSI on real or synthetic images and evaluated the performance of MSI detection in a held-out set of 142 patients. When trained on real images from training cohort 1, this system achieved an area under the receiver operating curve (AUROC) of 0.742 [0.681, 0.854]. Training on the larger cohort 2 only marginally improved the AUROC to 0.757 [0.707, 0.869]. Training on purely synthetic data resulted in an AUROC of 0.743 [0.658, 0.801]. Training on both real and synthetic data further increased AUROC to 0.777 [0.715, 0.821]. We conclude that synthetic histology images retain information reflecting underlying genetic alterations in colorectal cancer. Using synthetic instead of real images to train deep learning systems yields non-inferior classifiers. This approach can be used to create large shareable data sets or to augment small data sets with rare molecular features. [c] 2021 The Authors. The Journal of Pathology published by John Wiley & Sons, Ltd. on behalf of The Pathological Society of Great Britain and Ireland. Article Note: No conflicts of interest were declared. CAPTION(S): Figure S1. Visualization of the flow of all samples Figure S2. Additional activation maps, related to Figure 4 Table S1. Overview of the results Table S2. Overview of the image sets Table S3. Step-by-step explanation of experiment #1 in Table S1 Table S4. Hyperparameter sets Table S5. Step-by-step explanation of experiment #2 in Table S1 Byline: Jeremias Krause, Heike I Grabsch, Matthias Kloor, Michael Jendrusch, Amelie Echle, Roman David Buelow, Peter Boor, Tom Luedde, Titus J Brinker, Christian Trautwein, Alexander T Pearson, Philip Quirke, Josien Jenniskens, Kelly Offermans, Piet A Brandt, Jakob Nikolas Kather