An Automated Kidney Tumour Detection Technique from Computer Tomography Images

Abstract

The abnormal multiplication of cells in the kidney can lead to the formation of a tumor commonly known as kidney cancer. The early-stage diagnosis of tumors in the kidney can significantly improve the chances of recovery. There are several imaging techniques available to physicians to diagnose the stage of cancer and the response of the patients to the prescription. Various medical imaging techniques are extensively used for the diagnosis and detection of kidney tumors. Currently, the diagnosis and detection are the primary emphases of renal kidney cancer-related research besides recognizing whether the tumor is malignant or not. In this paper, CT images are utilized to spot and pinpoint tumor regions in kidneys using an image processing technique. The current image processing technique combines pre-processing, edge detection, and segmentation stages and anticipated a rapid diagnosis of tumor from CT scans. The available CT scan image can be transformed to a grayscale complement and subsequently subjected to noise reduction during the pre-processing stage. Various well-known algorithms are used in the second stage for detecting the edges. Finally, K means clustering and later on, K means segmentation is employed to distinguish the tumor grown region in the CT images of kidneys.