Michael Gadermayr

TL;DR: In this article, the authors focus on a particularly powerful class of architectures, the so-called generative adversarial networks (GANs) applied to histological image data, and present the main applications of GANs, and give an outlook of some chosen promising approaches and their possible future applications.

TL;DR: This paper develops approaches based on adversarial models for image-to-image translation relying on unpaired training for stain-independent supervised segmentation and develops a fully-unsupervised segmentation approach exploiting image- to- image translation to convert from the image to the label domain.

TL;DR: It is shown that especially one of the proposed cascade networks proved to be a highly powerful tool for segmenting the renal glomeruli providing best segmentation accuracies and also keeping the computing time at a low level.

TL;DR: Two different CNN cascade approaches are proposed which are subsequently applied to segment the glomeruli in whole slide images of the kidney and compared with conventional fully-convolutional networks, providing the best segmentation accuracies and also keeping the computing time at the lowest level.

TL;DR: This experimental study proposes and investigates two different pipelines for performing stain-independent segmentation of histological whole slide images requiring annotated training data for one single stain only, and provides evidence that the direction of translation plays a crucial role considering the final segmentation accuracy.