Stain-Independent Deep Learning-Based Analysis of Digital Kidney Histopathology.

Convolutional neural network (CNN)-based image analysis applications in digital pathology (eg, tissue segmentation) require a large amount of annotated data and are mostly trained and applicable on a single stain. Here, a novel concept based on stain augmentation is proposed to develop stain-independent CNNs requiring only one annotated stain. In this benchmark study on stain independence in digital pathology, this approach is comprehensively compared with state-of-the-art techniques including image registration and stain translation, and several modifications thereof. A previously developed CNN for segmentation of periodic acid-Schiff-stained kidney histology was used and applied to various immunohistochemical stainings. Stain augmentation showed very high performance in all evaluated stains and outperformed all other techniques in all structures and stains. Without the need for additional annotations, it enabled segmentation on immunohistochemical stainings with performance nearly comparable to that of the annotated periodic acid-Schiff stain and could further uphold performance on several held-out stains not seen during training. Herein, examples of how this framework can be applied for compartment-specific quantification of immunohistochemical stains for inflammation and fibrosis in animal models and patient biopsy specimens are presented. The results show that stain augmentation is a highly effective approach to enable stain-independent applications of deep-learning segmentation algorithms. This opens new possibilities for broad implementation in digital pathology.

Limitations of Nerve Fiber Density as a Prognostic Marker in Predicting Oncological Outcomes in Hepatocellular Carcinoma.

It has been shown that the presence and density of nerve fibers (NFs; NFD) in the tumor microenvironment (TME) may play an important prognostic role in predicting long-term oncological outcomes in various malignancies. However, the role of NFD in the prognosis of hepatocellular carcinoma (HCC) is yet to be explored. To this end, we aimed to investigate the impact of NFs on oncological outcomes in a large European single-center cohort of HCC patients. In total, 153 HCC patients who underwent partial hepatectomy in a curative-intent setting between 2010 and 2021 at our university hospital were included in this study. Group comparisons between patients with and without NFs were conducted and the association of recurrence-free survival (RFS) and overall survival (OS) with the presence of NFs and other clinico-pathological variables were determined by univariate and multivariable Cox regression models. Patients with NFs in the TME presented with a median OS of 66 months (95% CI: 30−102) compared to 42 months (95% CI: 20−63) for patients without NFs (p = 0.804 log-rank). Further, RFS was 26 months (95% CI: 12−40) for patients with NFs compared to 18 months (95% CI: 9−27) for patients without NFs (p = 0.666 log-rank). In a subgroup analysis, patients with NFD ≤ 5 showed a median OS of 54 months (95% CI: 11−97) compared to 48 months (95% CI: 0−106) for the group of patients with NFD > 5 (p = 0.787 log-rank). Correspondingly, the RFS was 26 months (95% CI: 10−42) in patients with NFD ≤ 5 and 29 months (95% CI: 14−44) for the subcohort with NFD > 5 (p = 0.421 log-rank). Further, group comparisons showed no clinico-pathological differences between patients with NFs (n = 76) and without NFs (n = 77) and NFs were not associated with OS (p = 0.806) and RFS (p = 0.322) in our Cox regression models. In contrast to observations in various malignancies, NFs in the TME and NFD are not associated with long-term oncological outcomes in HCC patients undergoing surgery.

Deep learning detects genetic alterations in cancer histology generated by adversarial networks.

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 10 000 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. © 2021 The Authors. The Journal of Pathology published by John Wiley & Sons, Ltd. on behalf of The Pathological Society of Great Britain and Ireland.