Machine Learning for Classification of Cutaneous Sebaceous Neoplasms: Implementing Decision Tree Model Using Cytological and Architectural Features

Abstract

Abstract Background: This observational study aimed to describe and compare histopathological, architectural, and nuclear characteristics of sebaceous lesions and utilized these characteristics to develop a predictive classification approach using machine learning algorithms. Methods: This cross-sectional study was conducted on patients with sebaceous from March 2015 to March 2019. Pathology slides were retrieved and reviewed. Two distinct pathologists assessed each slide regarding architectural and cytological attributes. A decision tree method was used to develop a prediction model. multiple models were trained on a random 80% train set, this time only using the selected variables, and mean accuracy was calculated. Results: This study assessed characteristics of 124 sebaceous tumors. Histopathological findings such as pagetoid appearance, neurovascular invasion, atypical mitosis, extensive necrotic area, poor cell differentiation, and non-lobular tumor growth pattern, as well as nuclear features such as highly irregular nuclear contour, and large nuclear size were exclusively observed in carcinomatous tumors. Among non-carcinomatous lesions, some sebaceoma cases had features like infiltrative tumor margin, and high mitotic activity which can be misleading and complicate diagnosis. Based on multiple decision tree models, the five most critical variables for lesion categorization were identified as: nuclear contour, nucleoli, peripheral basaloid cell layers, basaloid cell count, and chromatin. Conclusions: This study implemented a machine learning modeling approach to help categorize controversial sebaceous lesions based on architectural and nuclear features, optimally. However, studies of larger sample sizes are needed to ensure the accuracy of our suggested predictive model.