Digital Mammogram and Tumour Detection Using Fractal-Based Texture Analysis: A Box-Counting Algorithm

Abstract

Mammography and X-ray imaging of the breast are considered as the mainstay of breast cancer screening. In the past several years, there has been tremendous interest in image processing and analysis techniques in mammography. The fractal is an irregular geometric object with an infinite nesting of structure of different sizes. Fractals can be used to make models of any objects. The most important properties of fractals are self-similarity, chaos, and non-integer fractal dimension. The fractal dimension analysis has been applied to study the wide range of objects in biology and medicine and has been used to detect small tumors, microcalcification in mammograms, tumors in brain, and to diagnose blood cells and human cerebellum. Fractal theory also provides an appropriate platform to build oncological-related software program because the ducts within human breast tissue have fractal properties. Fractal analysis of mammogram was used for the breast parenchymal density assessment. The fractal dimension of the surface is determined by utilizing the Box-counting method. The Mammograms were collected from HCG Hospital, Bangalore. In this study, a method was developed in the Visual Basic for extracting the suspicious region from the mammogram based on texture. The fractal value obtained through Box-counting method for benign and malignant breast cancer is combined into a set. An algorithm was used to calculate the fractal value for the extracted image of the mammogram using Box-counting method.