English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Immunohistochemistry is a valuable diagnostic tool for cancer pathology.However, it requires specialist labs and equipment, is time-intensive, and isdifficult to reproduce. Consequently, a long term aim is to provide a digitalmethod of recreating physical immunohistochemical stains. GenerativeAdversarial Networks have become exceedingly advanced at mapping one image typeto another and have shown promise at inferring immunostains from haematoxylinand eosin. However, they have a substantial weakness when used with pathologyimages as they can fabricate structures that are not present in the originaldata. CycleGANs can mitigate invented tissue structures in pathology imagemapping but have a related disposition to generate areas of inaccuratestaining. In this paper, we describe a modification to the loss function of aCycleGAN to improve its mapping ability for pathology images by enforcingrealistic stain replication while retaining tissue structure. Our approachimproves upon others by considering structure and staining during modeltraining. We evaluated our network using the Fr\'echet Inception distance,coupled with a new technique that we propose to appraise the accuracy ofvirtual immunohistochemistry. This assesses the overlap between each staincomponent in the inferred and ground truth images through colour deconvolution,thresholding and the Sorensen-Dice coefficient. Our modified loss functionresulted in a Dice coefficient for the virtual stain of 0.78 compared with thereal AE1/AE3 slide. This was superior to the unaltered CycleGAN's score of0.74. Additionally, our loss function improved the Fr\'echet Inception distancefor the reconstruction to 74.54 from 76.47. We, therefore, describe an advancein virtual restaining that can extend to other immunostains and tumour typesand deliver reproducible, fast and readily accessible immunohistochemistryworldwide.

Ensuring accurate stain reproduction in deep generative networks for virtual immunohistochemistry