Detection of Pancreatic Cancer Using Clustering and Wavelet Transform Techniques

Pancreatic Cancer is the fifth tolling cancer disease in the United States and one of the top causes of cancer death in the western countries becoming a major public health problem worldwide. The radical surgical resection represents the only chance for cure but, unfortunately it is possible in only 15% of patients. Even at experienced centres the 5-year survival rates for the most favourable patients who undergo resection and adjuvant therapy are less than 20%. In this paper, a methodology is proposed for identifying the cancer affected part. Initially, the images are collected and then converted in to RGB as the original image is in greyscale format. K-means clustering technique is adopted so as to segment the images in order to capture the region of interest. Later, Haar wavelet transform is considered to compute the threshold values for the region of interest.