Mahdi Hemmasian Ettefagh - Member of the Department of Infrastructure Technologies of the National Cancer Society, Massachusetts, Boston, USA
Nahid Nafissi - Assistant professor, Iran Medical University, Tehran, Iran
Mojgan Kalantari - Assistant professor, Shahid-Beheshti University of Medical Sciences, Tehran, Iran
Nasrin Ahmadinejad - Associate Professor, Tehran University of Medical Sciences, Tehran, Iran

Introduction: Breast cancer is the most common cancer and the first cause of cancer death among women in the world. At present, mammography is the most appropriate method for diagnosis of breast cancer, but due to the type of breast tissue and the use of X-ray radiation with low density in the preparation of mammograms The images are low and, given the fact that the tumors have different sizes and shapes, the detection of lesions, especially in the early stages of tumor formation, is very difficult and tedious.Methods: In this study, images of 3217 patients referring to hospitals in the United States from 2003 to 2013 were investigated and an innovative method was implemented on the images and the result of the reproduction of images was referred to young and inexperienced radiologists for interpretation, and the result of the interpretation corresponding to the final diagnosis of pathology.Findings: The new proposed method can increase the accuracy of detection in mammography images in less experienced radiologists up to 86.7% .The patent number 91836 on the 4/15/2017 was registered at the Iranian Intellectual Property Registration Office.Conclusions: By categorizing the optical spectrum in mammogram images based on the proposed method, the accuracy of detection in mammographic images can be increased to 86.7%.

Mammography, Breast Cancer, Computer Aided Diagnosis, Early stage