Using light microscopy to describe the microarchitecture of normal and diseased tissues has changed very little since the middle of the 19th century. While the premise of histologic analysis remains intact, our relationship with the microscope is changing dramatically. Digital pathology offers new forms of visualization, and delivery of images is facilitated in unprecedented ways. This new technology can untether us entirely from our light microscopes, with many pathologists already performing their jobs using virtual microscopy. Several veterinary colleges have integrated virtual microscopy in their curriculum, and some diagnostic histopathology labs are switching to virtual microscopy as their main tool for the assessment of histologic specimens. Considering recent technical advancements of slide scanner and viewing software, digital pathology should now be considered a serious alternative to traditional light microscopy. This review therefore intends to give an overview of the current digital pathology technologies and their potential in all fields of veterinary pathology (ie, research, diagnostic service, and education). A future integration of digital pathology in the veterinary pathologist's workflow seems to be inevitable, and therefore it is proposed that trainees should be taught in digital pathology to keep up with the unavoidable digitization of the profession.

We introduce a novel, large-scale dataset for microscopy cell annotations. The dataset includes 32 whole slide images (WSI) of canine cutaneous mast cell tumors, selected to include both low grade cases as well as high grade cases. The slides have been completely annotated for mitotic figures and we provide secondary annotations for neoplastic mast cells, inflammatory granulocytes, and mitotic figure look-alikes. Additionally to a blinded two-expert manual annotation with consensus, we provide an algorithm-aided dataset, where potentially missed mitotic figures were detected by a deep neural network and subsequently assessed by two human experts. We included 262,481 annotations in total, out of which 44,880 represent mitotic figures. For algorithmic validation, we used a customized RetinaNet approach, followed by a cell classification network. We find F1-Scores of 0.786 and 0.820 for the manually labelled and the algorithm-aided dataset, respectively. The dataset provides, for the first time, WSIs completely annotated for mitotic figures and thus enables assessment of mitosis detection algorithms on complete WSIs as well as region of interest detection algorithms.

Integration of new technologies, such as digital microscopy, into a highly standardized laboratory routine requires the validation of its performance in terms of reliability, specificity, and sensitivity. However, a validation study of digital microscopy is currently lacking in veterinary pathology. The aim of the current study was to validate the usability of digital microscopy in terms of diagnostic accuracy, speed, and confidence for diagnosing and differentiating common canine cutaneous tumor types and to compare it to classical light microscopy. Therefore, 80 histologic sections including 17 different skin tumor types were examined twice as glass slides and twice as digital whole-slide images by 6 pathologists with different levels of experience at 4 time points. Comparison of both methods found digital microscopy to be noninferior for differentiating individual tumor types within the category epithelial and mesenchymal tumors, but diagnostic concordance was slightly lower for differentiating individual round cell tumor types by digital microscopy. In addition, digital microscopy was associated with significantly shorter diagnostic time, but diagnostic confidence was lower and technical quality was considered inferior for whole-slide images compared with glass slides. Of note, diagnostic performance for whole-slide images scanned at 200× magnification was noninferior in diagnostic performance for slides scanned at 400×. In conclusion, digital microscopy differs only minimally from light microscopy in few aspects of diagnostic performance and overall appears adequate for the diagnosis of individual canine cutaneous tumors with minor limitations for differentiating individual round cell tumor types and grading of mast cell tumors.

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 intrarater variability, we propose our automated object detection pipeline, enabling accurate, reproducible and quick EIPH scoring in WSI. Patients with pulmonary hemorrhage (P-Hem) suffer from repeated bleeding into the lungs, which can result in dyspnea and if untreated, may have life threatening consequences 1. There are various causes which lead to P-Hem, including drug abuse, premature birth, leukaemia, autoimmune disorders and immunodeficiencies 2-6. In this paper, we focus on a special subtype of P-Hem called exercise-induced pulmonary hemorrhage (EIPH) in horses. Although EIPH also affects healthy human athletes 7 and racing greyhounds 8 , it is diagnosed most commonly in racing horses and causes reduced athletic performance 9-12. The gold standard for diagnosis of P-Hem in humans and equine animals is to perform cytology of bronchoalveolar lavage fluid (BALF) 4,13 using a scoring system as explained by Golde et al. 4. The red blood cells of the bleeding are degraded into an iron-storage complex called hemosiderin by alveolar macrophages. Hemosiderin-laden macrophages are called hemosiderophages. Prior to microscopic evaluation, the cells are extracted by the BALF procedure and stained with Perlss' Prussian Blue 14 or Turnbull's Blue 15 in order to visualise the iron pigments contained in the hemosiderin. According to the commonly used scoring system (macrophages hemosiderin score) by Golde et al. 4 , alveolar macrophages can be distinguished into five grades depending on their hemosiderin content. This scoring system is based on the principle that a higher score correlates with increased alveolar bleeding 16 .

Manual count of mitotic figures, which is determined in the tumor region with the highest mitotic activity, is a key parameter of most tumor grading schemes. It can be, however, strongly dependent on the area selection due to uneven mitotic figure distribution in the tumor section. We aimed to assess the question, how significantly the area selection could impact the mitotic count, which has a known high inter-rater disagreement. On a data set of 32 whole slide images of H&E-stained canine cutaneous mast cell tumor, fully annotated for mitotic figures, we asked eight veterinary pathologists (five board-certified, three in training) to select a field of interest for the mitotic count. To assess the potential difference on the mitotic count, we compared the mitotic count of the selected regions to the overall distribution on the slide. Additionally, we evaluated three deep learning-based methods for the assessment of highest mitotic density: In one approach, the model would directly try to predict the mitotic count for the presented image patches as a regression task. The second method aims at deriving a segmentation mask for mitotic figures, which is then used to obtain a mitotic density. Finally, we evaluated a two-stage object-detection pipeline based on state-of-the-art architectures to identify individual mitotic figures. We found that the predictions by all models were, on average, better than those of the experts. The two-stage object detector performed best and outperformed most of the human pathologists on the majority of tumor cases. The correlation between the predicted and the ground truth mitotic count was also best for this approach (0.963–0.979). Further, we found considerable differences in position selection between pathologists, which could partially explain the high variance that has been reported for the manual mitotic count. To achieve better inter-rater agreement, we propose to use a computer-based area selection for support of the pathologist in the manual mitotic count.