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 Plus annual

Presents the access of premium content as premium feature

Premium Content

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 Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Most tumor grading systems for human as for veterinary histopathology arebased upon the absolute count of mitotic figures in a certain reference area ofa histology slide. Since time for prognostication is limited in a diagnosticsetting, the pathologist will often almost arbitrarily choose a certain fieldof interest assumed to have the highest mitotic activity. However, as mitoticfigures are commonly very sparse on the slide and often have a patchydistribution, this poses a sampling problem which is known to be able toinfluence the tumor prognostication. On the other hand, automatic detection ofmitotic figures can't yet be considered reliable enough for clinicalapplication. In order to aid the work of the human expert and at the same timereduce variance in tumor grading, it is beneficial to assess the whole slideimage (WSI) for the highest mitotic activity and use this as a reference regionfor human counting. For this task, we compare two methods for region ofinterest proposal, both based on convolutional neural networks (CNN). For bothapproaches, the CNN performs a segmentation of the WSI to assess mitoticactivity. The first method performs a segmentation at the original imageresolution, while the second approach performs a segmentation operation at asignificantly reduced resolution, cutting down on processing complexity. Weevaluate the approach using a dataset of 32 completely annotated whole slideimages of canine mast cell tumors, where 22 were used for training of thenetwork and 10 for test. Our results indicate that, while the overallcorrelation to the ground truth mitotic activity is considerably higher (0.94vs. 0.83) for the approach based upon the fine resolution network, the field ofinterest choices are only marginally better. Both approaches propose fields ofinterest that contain a mitotic count in the upper quartile of respectiveslides.

Field Of Interest Proposal for Augmented Mitotic Cell Count: Comparison of two Convolutional Networks