A Deep Learning Approach for the Segmentation of Kidney, Tumor and Cyst in Computed Tomography Scans

Kidney cancer, also known as renal cell carcinoma, is a malignant tumor that originates in the kidneys. It is one of the most common types of cancer affecting the urinary system. Kidney tumors can vary in size, location, and aggressiveness, making early detection and accurate diagnosis crucial for effective treatment planning. The proposed method makes use of nnU-Net which is a self-adapting semantic segmentation method, to segment the kidney, tumor and cyst. The proposed neural network model was trained using the datasets provided by the 2023 Kidney and Kidney Tumor Segmentation Challenge hosted by MICCAI 2023 conference. The proposed methodology leveraged the power of deep learning to yield high segmentation accuracy.