Sessions

Hemangiosarcoma (HSA), the malignancy of vascular endothelium is a common splenic tumour in dogs which is characterized by poor prognosis, high mortality rates and frequent metastasis. Tumour development is frequently asymptomatic, and the late diagnosis limits the possibility of early intervention. Tumour grading is the basis for treatment plans and prognosis, but splenic HSA grading is not routinely used in dogs, partially due to its subjectivity. This study aimed to digitally quantify the extent of vascular formations and necrotic areas in canine splenic HSAs and to use these measurements to determine a digitally corrected grade which can be further compared to the manual grading. 38 histological slides of splenic HSAs were digitized, and tumour regions were examined by QuPath digital image analysis. 28 out of the 38 slides (74%) received the same score with a Fleiss kappa value of 0.486, which shows only moderate agreement. The lack of grade 3 tumours, the author’s experience in pathology and classifier training and the limitations of the software could be responsible for this finding. Despite these limitations, partially digitized grading systems can potentially improve objectivity, reproducibility and shorten examination time. For digital pathology to be implemented into grading systems effectively, expert pathologists need to work closely with image analysis to ensure accurate results.