Detection of Necrosis in Digitised Whole-Slide Images for Better Grading of Canine Soft-Tissue Sarcomas Using Machine-Learning

SIMPLE SUMMARY: Canine soft-tissue sarcomas are a group of tumours that arise from the skin and subcutaneous connective tissue. The most common method used to predict the behaviour of these tumours is grading. The grading system used for soft-tissue sarcomas is derived from a combined score calculated by evaluating the mitotic count, percentage of tumour necrosis and degree of cellular differentiation. However, these parameters are highly subjective and a high inter-observer variability has been reported in grading these tumours, which can result in complications regarding treatment plans. Manual identification of areas of necrosis is a time-consuming task that is prone to observer error. Artificial-intelligence algorithms and, in particular, machine learning, can help improve grading by automatically detecting regions of necrosis. The aim of this study was to differentiate image regions in order to automatically identify tumour necrosis in digitised canine soft-tissue sarcoma slides. This method showed an accuracy of 92.7% which represents the number of correctly classified data instances over the total number of data instances. Therefore, the proposed method is a promising tool to minimise human error in the evaluation of necrosis in soft-tissue sarcomas, and hence increase the efficiency and accuracy of histopathological grading of canine soft-tissue sarcomas. ABSTRACT: The definitive diagnosis of canine soft-tissue sarcomas (STSs) is based on histological assessment of formalin-fixed tissues. Assessment of parameters, such as degree of differentiation, necrosis score and mitotic score, give rise to a final tumour grade, which is important in determining prognosis and subsequent treatment modalities. However, grading discrepancies are reported to occur in human and canine STSs, which can result in complications regarding treatment plans. The introduction of digital pathology has the potential to help improve STS grading via automated determination of the presence and extent of necrosis. The detected necrotic regions can be factored in the grading scheme or excluded before analysing the remaining tissue. Here we describe a method to detect tumour necrosis in histopathological whole-slide images (WSIs) of STSs using machine learning. Annotated areas of necrosis were extracted from WSIs and the patches containing necrotic tissue fed into a pre-trained DenseNet161 convolutional neural network (CNN) for training, testing and validation. The proposed CNN architecture reported favourable results, with an overall validation accuracy of 92.7% for necrosis detection which represents the number of correctly classified data instances over the total number of data instances. The proposed method, when vigorously validated represents a promising tool to assist pathologists in evaluating necrosis in canine STS tumours, by increasing efficiency, accuracy and reducing inter-rater variation.