Quantitative analysis of pre-processing techniques for tumour detection

Cancer is a group of diseases characterized by uncontrolled growth and spread of abnormal cells. This paper aims at to detect the tumour region in the MRI and segment it. The research involved here investigates different filtering techniques for pre processing which include Gaussian filter, Median filter, Order statistic filter and Wiener filter. This paper further analyzes segmentation techniques such GLCM based segmentation which extracts features from overlapping blocks and the classification of the tumorous region is done by k-means clustering. Different cluster sizes and pre-processing of the extracted blocks for the GLCM based technique are compared based on efficiency parameters such as accuracy and tumour detection percentage.