RESUMEN
Microscopic evaluation of hematoxylin and eosin-stained slides is still the diagnostic gold standard for a variety of diseases, including neoplasms. Nevertheless, intra- and interrater variability are well documented among pathologists. So far, computer assistance via automated image analysis has shown potential to support pathologists in improving accuracy and reproducibility of quantitative tasks. In this proof of principle study, we describe a machine-learning-based algorithm for the automated diagnosis of 7 of the most common canine skin tumors: trichoblastoma, squamous cell carcinoma, peripheral nerve sheath tumor, melanoma, histiocytoma, mast cell tumor, and plasmacytoma. We selected, digitized, and annotated 350 hematoxylin and eosin-stained slides (50 per tumor type) to create a database divided into training, n = 245 whole-slide images (WSIs), validation (n = 35 WSIs), and test sets (n = 70 WSIs). Full annotations included the 7 tumor classes and 6 normal skin structures. The data set was used to train a convolutional neural network (CNN) for the automatic segmentation of tumor and nontumor classes. Subsequently, the detected tumor regions were classified patch-wise into 1 of the 7 tumor classes. A majority of patches-approach led to a tumor classification accuracy of the network on the slide-level of 95% (133/140 WSIs), with a patch-level precision of 85%. The same 140 WSIs were provided to 6 experienced pathologists for diagnosis, who achieved a similar slide-level accuracy of 98% (137/140 correct majority votes). Our results highlight the feasibility of artificial intelligence-based methods as a support tool in diagnostic oncologic pathology with future applications in other species and tumor types.
Asunto(s)
Aprendizaje Profundo , Enfermedades de los Perros , Neoplasias Cutáneas , Animales , Perros , Inteligencia Artificial , Eosina Amarillenta-(YS) , Hematoxilina , Reproducibilidad de los Resultados , Neoplasias Cutáneas/diagnóstico , Neoplasias Cutáneas/veterinaria , Aprendizaje Automático , Enfermedades de los Perros/diagnósticoRESUMEN
BACKGROUND: Even though tumours are considered to occur frequently in pet hamsters, there is only a small number of scientific reports in current literature. METHODS: Pathological reports from 177 hamsters were evaluated. RESULTS: Of these, 78 were male and 75 were female. Median age of affected hamsters was 12 months (range 2-34). Integumental tumours were the most common neoplasms (62%, 109/177). As far as species was known, the number of Syrian hamsters (52%, 30/58) affected by tumours seemed to be lower than the number of affected dwarf hamsters (85%, 47/55). Tumours of the hematopoietic system were the second most frequently neoplasms (17%, 30/177). Relative number of neoplasms of the endocrine system, tumours of the digestive system (1.7%, 3/177) and other tumours (4%, 7/177 each) was low. The majority of integumental tumours were epithelial (66%; 91/126). CONCLUSION: This study aimed to analyze data from veterinary surgeries and pathological institutes about the occurrence of spontaneous tumours in Syrian hamsters and dwarf hamsters to give information about the frequency of tumours, prognosis and survival time. This is the first study about tumours in pet hamsters in Germany so far.
Asunto(s)
Neoplasias/veterinaria , Mascotas , Enfermedades de los Roedores/epidemiología , Animales , Cricetinae , Femenino , Alemania/epidemiología , Masculino , Neoplasias/epidemiología , Estudios RetrospectivosRESUMEN
Prevalence and age distribution of tumors is largely unknown in pet rabbits. Currently available studies focused on specific organ systems or specific tumor types and never covered a comparative examination of all tumor types. Previous studies on laboratory rabbits suggested a low tumor prevalence but were mostly limited to young adult animals. In the present study, all tumor types and several tumor-like lesions of all organ systems were analyzed retrospectively in archived pet rabbit samples of all ages. Cases included necropsy cases (n = 2,014) or postmortem tissue samples (n = 102) as well as surgical biopsies (n = 854). All lesions suspicious of neoplasia were reevaluated by histopathology and, when indicated, by immunohistochemistry. Necropsy cases had a tumor prevalence of 14.4% in both sexes or 19.8% in female intact rabbits of all age groups, and up to 47.2% or 66.7%, respectively, in rabbits older than 6 years. Overall, the most common tumor types were uterine adenocarcinoma (prevalence in female intact animals: 13.1%), lymphoma (prevalence: 2.8%), and thymoma (prevalence: 2.1%). Lymphoma, the most common tumor of rabbits ≤24 months of age, were of B-cell immunophenotype in 96% of cases and most commonly located in the lymph nodes (57%), gastrointestinal tract (54%), kidneys (48%), spleen (42%), and liver (41%). Tumors accounted for 81.1% of surgical biopsies and mostly comprised cutaneous, mammary, and uterine tumors. In conclusion, tumor types and prevalence varied significantly with respect to age, revealing some differences from previous studies on laboratory rabbits.
Asunto(s)
Linfoma , Neoplasias Uterinas , Animales , Femenino , Inmunohistoquímica , Inmunofenotipificación/veterinaria , Linfoma/veterinaria , Masculino , Conejos , Estudios Retrospectivos , Neoplasias Uterinas/veterinariaRESUMEN
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.