Artificial Neural Network based Detection of Renal Tumors using CT Scan Image Processing

摘要

Renal tumour segmentation and analysis is a very important step for doctors in deciding the stage of cancer and determining the method of treatment. This paper examines a novel approach to develop an efficient algorithm to detect and further analyse the renal cancer tumours. The algorithm has been employed to pre-process and segment the image for better visualization and segmentation of the visible tumour. The pre-processing involves hybrid filter for noise removal and image enhancement. An artificial neural network has also been used by means of Hybrid Self Organizing Maps using which we have used for clustering of the image data and thereby highlighting the detected region. The correct output obtained by the medical team is then compared with the resultant image in order to improve algorithm to aptly understand the affected regions in human body and aid in better visualization of the tumor. We then apply a region growing method which looks for similar intensity regions in the images and thus segment outs the tumour from the processed image.