Indian Journal of Science and Technology
Year: 2015, Volume: 8, Issue: 26, Pages: 1-7
S. L. Aarthy* and S. Prabu
Background/Objectives: The goal of screening exams for breast cancer is to find cancer before they start to cause symptoms. The objective of this paper is to propose a diagnosis tool to detect breast cancer at the earlier stage. Methods/ Statistical Analysis: The proposed diagnosis tool for detecting breast cancer involves the various processes which includes wavelets. The diagnosis tool is proposed to execute multiscalar contrast enhancement at different wavelet scales. The process is continued with segmentation techniques which combine information of the core image structure and the boundaries. The approach confirms the use of a single algorithm for image processing and detects both microclassification and masses in the breast image. Findings: An algorithm has been proposed for the enhancement and denoising of an image with the help of processing through wavelets. Moreover, in case of microcalcification, a general method is described. The function is to execute multiscalar contrast enhancement at different wavelets scales. On the other hand mass detection means a segmentation of image where mathematical morphology is mixed with dyadic wavelet information. Segmentation technique is divided in two different groups: 1. Discrete contour model and 2. Region growing. The proposed system was able to identify the masses and micro classification without any biopsy and other process. In this paper we have proposed a technique that introduces segmentation combining efficiently both the information of the core of image structure and the boundaries. This approach confirms the use of a single algorithm for image processing and detects both microcalcification and masses. Application/Improvements: The proposed diagnosis tool is also used for enhancing tiny features which can be still accurate in cancer detection. Further the algorithm can be used to classify the breast image as malignant or benign.
Keywords: Enhancement, Masses, Matlab, Microclasification, Morphology, Region Growing, Wavelet
Subscribe now for latest articles and news.