Chemotherapy in cancer patients may induce damage to the renal vascular system. Accurate segmentation and analysis of renal vascular structure play a pivotal role in early diagnosis and treatment planning for kidney diseases. In this paper, we propose a Renal Damage Detection System (RDDS) for the early diagnosis of kidney damage. It features VoxUnet, a custom 3D segmentation model that offers enhanced performance in delineating intricate vascular structures from medical imaging data with an accuracy of 99.2% and a dice score of 92.5%. When the renal vascular structure is damaged, it can narrow or block blood vessels in the kidneys. This reduces blood flow to the kidneys, hindering their ability to filter waste products from the blood. A 3D model is built from the segmented images, which is used to detect damage in the renal vascular system. Utilizing a Variational Autoencoder with FoldingNet, the model successfully identifies ruptures and structural abnormalities within the blood vessels, achieving an overall Area Under the Curve (AUC) of 76.3% in anomaly detection.