Automating Kidney and Kidney Tumor Segmentation Using Deep Learning
Carlos De Leon, Nathalie Aquino, Gwenndolyn Campbell
Department of Computer Science
Faculty Supervisor: Sara El Alaoui
Kidney tumors, whether benign or malignant, require early detection in order to be effectively treated, and often rely on medical images such as CT scans. Manually analyzing these images can be very time consuming and can heavily prolong the process of an official diagnosis and treatment. In this study, we utilize the 2019 Kidney and Kidney Tumor Segmentation (KiTS19) dataset, which includes 300 kidney tumor cases with CT images and segmentation labels to develop a deep learning model for automated kidney and tumor segmentation. Using convolutional neural networks (CNNs), we aim to create an efficient pipeline for accurately identifying and segmenting kidney tissues and tumors from three-dimensional CT scans. This interdisciplinary project bridges biomedical science and computer science, showing the potential of deep learning in advancing medical image analysis.