Texture and statistical analysis of mammograms: A novel method to detect tumor in Breast Cells

Abstract

There are countless ways the human body fails. Breast cancer is one of them, especially for women. It is the most common cancer of women worldwide. It has been reported by the US Breast Cancer Registry that more than 25% and up to 50% of the decline in mortality was due to the increased use of screening mammography. The detection accuracy of these mammograms could be enhanced using suitable numerical algorithms, to reduce the amount of false positives and negatives, which are 20% and 10% respectively. We have used sophisticated texture and statistical feature extraction algorithms to increase the accuracy up to 98%. The texture technique is more robust than the statistical analysis. These methods have the potential to transfer to clinic as well as to use as mobile apps for a second opinion.