Breast cancer is the most common type of cancer among women. The International Agency for Research on Cancer (IARC) has estimated that more than one million cases of breast cancer occur annually in the world. Early detection of breast cancer plays an important role in reducing the mortality rate and makes this type of cancer one of the most treatable cancers. Due to the ability to detect tumors and microcalcifications in the early stages of cancer, mammography is recommended every three years for all women between 50 and 95 years old. A large number of mammography images on the one hand and the dependence of the diagnosis of calcium clusters on the doctor’s experience on the other hand have led researchers to use computer-aided diagnosis systems. Due to the small size of calcareous deposits in mammography images, the analysis of these images requires a lot of accuracies, and the complete analysis of an image causes fatigue and reduces the doctor’s accuracy. Therefore, designing an automatic system to identify areas with lesions will increase the accuracy of the radiologist’s diagnosis and increase the speed of screening images.
Screening through mammography has a great effect on reducing the death rate from breast cancer, and screening programs have reduced the death rate by 35-75%.
In recent years, artificial intelligence (AI) has played a prominent role in the examination of mammography images as well as the diagnosis and classification of lesions present in these images. In AI-based methods, the pixel characteristics of the whole image are extracted, and based on that, the presence or absence of fine
Calcareous deposits or Micro Calcifications are detected. MCFinder software system is one of the best tools for detecting these lesions in mammography images.
The powerful MCFinder software has been developed to process mammography images and detect benign and malignant Micro Calcifications to better screen patients and different people. In this software, by processing a mammography image, the area of the presence of fine limestone particles, if any, is determined and shown. The images related to the software are shown in the figures below. To evaluate the accuracy of this software, 152 mammography images of 76 people, including 40 patients and 36 healthy cases, from the DDSM database have been processed by this software. The obtained results show that out of 40 positive cases, only three false negative cases were reported, and also only five false positive cases were reported out of 36 negative cases. Therefore, the detection accuracy of the software is 90% and its sensitivity is 92.5%. In addition, the processing time of each image by this software is about 4 minutes on average.
