Contrast Enhancement Of Mammograms For Breast Cancer Detection

Abstract

Breast cancer is known to be the most common form of cancer in women. Every 1 in 10 woman is diagnosed with the same. It has become one of the main diseases according to published statistics over the past 50 years. Until now, however, it is not considered that there have been any useful ways to avoid breast cancer, as its causes are still unknown. Studies have shown its survival has improved in the recent years as early detection and appropriate treatment of breast cancer significantly increased the chances to live. The typical diagnostic signs such as microcalcifications and masses are difficult to detect because mammograms are low-contrast and noisy images and can’t prove that an abnormal area is cancer. Therefore, image denoising and enhancement becomes necessary for better understanding of stereotypes. This project reviews the recent enhancement techniques that have been explored and exist in the literature. At the enhancement stage, the geometric information is analyzed to increase the contrast of the edges. Denoising and enhancement stages can be applied together for the enhancement of mammogram images in various ways. Each method is classified, analyzed, and compared against other approaches. To examine the accuracy of the mammogram enhancement techniques, the sensitivity and specificity of the approaches is presented and compared where applicable. This project will provide anatomy for the available approaches and highlights the best available enhanc