G-008
Automatic Mass Detection for Breast Cancer from Mammography with Texture Analysis Technique and Statistical features

◎Somchanok Tivatansakul・Keiichi Uchimura（Kumamoto Univ.）

This study focuses on precise detection of mass boundary from mammography. We apply gray-level co-occurrence matrix (GLCM) with statistical features. We also improve it using preprocessing and GLCM iteration to distinguish, remove and detect mass regions from other breast areas. The evaluation results with a mini-MIAS database of mammograms (MIAS) indicate that the method is more suitable for detection of well-defined, circumscribed, ill-defined and other mass types. However, it has some difficulty to accurately detect any masses in high dense breast area with unclear margin. This case will be taken into account as our future works.