| Autor: |
Marzahl C; Pattern Recognition Lab, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany. c.marzahl@euroimmun.de.; Research and Development, EUROIMMUN Medizinische Labordiagnostika AG, Lübeck, Germany. c.marzahl@euroimmun.de., Aubreville M; Pattern Recognition Lab, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany., Bertram CA; Institute of Veterinary Pathology, Freie Universität Berlin, Berlin, Germany., Stayt J; VetPath Laboratory Services, Ascot, Western, Australia., Jasensky AK; Laboklin GmbH und Co. KG, Bad Kissingen, Germany., Bartenschlager F; Institute of Veterinary Pathology, Freie Universität Berlin, Berlin, Germany., Fragoso-Garcia M; Institute of Veterinary Pathology, Freie Universität Berlin, Berlin, Germany., Barton AK; Equine Clinic, Freie Universität Berlin, Berlin, Germany., Elsemann S; Department of Neurosurgery, Universitätsklinikum Erlangen, Erlangen, Germany., Jabari S; Institute of Neuropathology, Friedrich Alexander University Erlangen-Nürnberg, Erlangen, Germany., Krauth J; Research and Development, EUROIMMUN Medizinische Labordiagnostika AG, Lübeck, Germany., Madhu P; Pattern Recognition Lab, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany., Voigt J; Research and Development, EUROIMMUN Medizinische Labordiagnostika AG, Lübeck, Germany., Hill J; VetPath Laboratory Services, Ascot, Western, Australia., Klopfleisch R; Institute of Veterinary Pathology, Freie Universität Berlin, Berlin, Germany., Maier A; Pattern Recognition Lab, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany. |
| Abstrakt: |
Exercise-induced pulmonary hemorrhage (EIPH) is a common condition in sport horses with negative impact on performance. Cytology of bronchoalveolar lavage fluid by use of a scoring system is considered the most sensitive diagnostic method. Macrophages are classified depending on the degree of cytoplasmic hemosiderin content. The current gold standard is manual grading, which is however monotonous and time-consuming. We evaluated state-of-the-art deep learning-based methods for single cell macrophage classification and compared them against the performance of nine cytology experts and evaluated inter- and intra-observer variability. Additionally, we evaluated object detection methods on a novel data set of 17 completely annotated cytology whole slide images (WSI) containing 78,047 hemosiderophages. Our deep learning-based approach reached a concordance of 0.85, partially exceeding human expert concordance (0.68 to 0.86, mean of 0.73, SD of 0.04). Intra-observer variability was high (0.68 to 0.88) and inter-observer concordance was moderate (Fleiss' kappa = 0.67). Our object detection approach has a mean average precision of 0.66 over the five classes from the whole slide gigapixel image and a computation time of below two minutes. To mitigate the high inter- and intra-rater variability, we propose our automated object detection pipeline, enabling accurate, reproducible and quick EIPH scoring in WSI. |
