ساحل غيبي پور

After skin cancer, breast cancer is the most common cancer among women and the second cause of death from cancer, after lung cancer. The incidence of this disease, which increases every year, leads to the idea that this cancer cannot be cured. Early detection of breast cancer can be very useful to minimize the destructive effects of this disease on sufferers. This research proposes a classification for breast ommography detection using probabilistic neural network (PNN) as a classifier, we use median filters to minimize noise to optimize the pre-processing steps. For segmentation, we use Gaussian mixture model (GMM) and shell game optimization, and for feature extraction we use gray covariance matrix method (GLCM). The PNN classifier was chosen because it can work efficiently on small datasets. The effectiveness of the proposed method was eva‎luated by eva‎luating its accuracy and sensitivity criteria in the Mias data set. The results show that the proposed method has significant performance and is very accurate in detecting tumors of different sizes in complex and high-resolution images. In addition, our method outperforms existing diagnostic methods.

سرطان سينه , تصاويرماموگرافي , تشخيص توده , شبكه عصبي , بخش بندي

breast cancer , mammography images , Mass diagnosis , neural network , Segmentation

Sahel Gheibipoor

Dr. Rahman Farnoosh