Portal de Periódicos da CAPES

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

2022; Elsevier BV; Volume: 193; Issue: 1 Linguagem: Inglês

10.1016/j.ajpath.2022.09.011

1525-2191

Nassim Bouteldja, David L. Hölscher, Barbara M. Klinkhammer, Roman D. Buelow, Johannes Lotz, Nick Weiss, Christoph Daniel, Kerstin Amann, Peter Boor,

Generative Adversarial Networks and Image Synthesis

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. 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. Creating a More Welcoming Home for Your Work at The American Journal of PathologyThe American Journal of PathologyVol. 193Issue 1PreviewWhen Martha Furie became Editor-in-Chief of The American Journal of Pathology (AJP) in 2018, her target was for AJP to achieve an Impact Factor of greater than 5.0 by the end of her five-year term. We were delighted to learn this past June that we made it! The most recent Impact Factor is 5.77, and AJP remains the most highly cited journal in the field of pathology. This goal was attained through the dedicated efforts of the AJP team; the contribution of high-quality, original research by AJP's authors; and an emphasis on increasing the number of timely review articles and theme issues. Full-Text PDF