Health AI Central
Share:
Peer-Reviewed Publication
Sci Rep2024;14(1):26273.November 1, 2024Journal Article

Information mismatch in PHH3-assisted mitosis annotation leads to interpretation shifts in H&E slide analysis.

Jonathan Ganz1, Christian Marzahl2, Jonas Ammeling1, Emely Rosbach1, Barbara Richter3, Chloé Puget4, Daniela Denk5,6, Elena A Demeter7, Flaviu A Tăbăran8, Gabriel Wasinger9, Karoline Lipnik3, Marco Tecilla10, Matthew J Valentine11, Michael J Dark12, Niklas Abele13, Pompei Bolfa11, Ramona Erber13,14, Robert Klopfleisch4, Sophie Merz15, Taryn A Donovan16, Samir Jabari13, Christof A Bertram3, Katharina Breininger17,18, Marc Aubreville19,20
1Technische Hochschule Ingolstadt, Ingolstadt, Germany.
2Gestalt Diagnostics, Spokane, USA.
3University of Veterinary Medicine, Vienna, Austria.
4Freie Universität Berlin, Berlin, Germany.
5Ludwig-Maximilians-Universität München, München, Germany.
6SeaWorld Yas Island, Abu Dhabi, UAE.
7Cornell University, Ithaca, USA.
8University of Agricultural Sciences and Veterinary Medicine of Cluj-Napoca, Cluj-Napoca, Romania.
9Medical University of Vienna, Vienna, Austria.
10University of Milan, Milan, Italy.
11Ross University School of Veterinary Medicine, Basseterre, St. Kitts and Nevis.
12University of Florida, Gainesville, USA.
13University Hospital Erlangen, Erlangen, Germany.
14Universität Regensburg, Regensburg, Germany.
15IDEXX Laboratories, Kornwestheim, Germany.
16The Schwarzman Animal Medical Center, New York, USA.
17Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany.
18Universität Würzburg, Würzburg, Germany.
19Technische Hochschule Ingolstadt, Ingolstadt, Germany. marc.aubreville@hs-flensburg.de.
20Flensburg University of Applied Sciences, Flensburg, Germany. marc.aubreville@hs-flensburg.de.
PubMed PMC DOI PDF

Abstract

The count of mitotic figures (MFs) observed in hematoxylin and eosin (H&E)-stained slides is an important prognostic marker, as it is a measure for tumor cell proliferation. However, the identification of MFs has a known low inter-rater agreement. In a computer-aided setting, deep learning algorithms can help to mitigate this, but they require large amounts of annotated data for training and valid…

Create a free account to keep reading

Free members get 10 full research views every month across publications, clinical trials, FDA clearances, adverse events, and NIH grants. No credit card required.

Create Free AccountSign In

Want unlimited research access? See Pro plans

Data Accuracy Notice: Research intelligence on Health AI Central is aggregated from public sources (PubMed, ClinicalTrials.gov, FDA, NIH, CMS, and others) and refreshed nightly. Classifications and derived metrics are produced by automated methods described in our Methodology. We recommend verifying critical data points against the primary sources before making decisions.