Classification of Microcalcification Areas Based on Extraction Of Local Features From Breast Ultrasound images

mahshid hemati nik¹ nader jafarnia dabanloo² elham keshavarz³ ensie khalili⁴

1) 1. Department of Engineering of biomedical engineering, Science and Research University, Tehran, Iran
2) 1. Department of Engineering of biomedical engineering, Science and Research University, Tehran, Iran
3) 1. Department of Engineering of biomedical engineering, shahid beheshti University, Tehran, Iran
4) 1. Department of Engineering of biomedical engineering, shahid beheshti University, Tehran, Iran

Publication : 2nd. International Congress on science & Engineering - paris(parisconf.com)
Abstract :
Calcification of breast Calcium sedimentation in breast tissue. In mamography, these sediments are seen in the form of white dots or lines, and are usually small enough that they cannot be felt. Breast calcifications are usually visible in mamography and it is much easier to diagnose them after menopause. Although these calcifications are usually non-cancerous (benign), some specific breast-calcification patterns-such as clusters of clusters or abnormal forms-can be indicative of breast cancer. In mammograms, calcifications are seen as micro-calcification or macro-calcification. Macro calcifications appear in the form of large white lines. In the vast majority of cases, macro-calcifications are non-cancerous and do not require any supplementary therapies or therapies. Micro-calcifications appear in the form of small white spots or spots, much like salt grains. Micro calcifications are often benign, but some calcification patterns can also indicate cancer. if the calcification pattern is suspected, the radiologist will do more tests, such as mammography with enlarged images or chest or ultrasound imaging. The best way to detect cancer and lack of tissue is through the use of sampling and then assessing that the problem with this method is that it is invasive. Mammography is currently the most appropriate diagnostic method for breast cancer. But due to the type of breast tissue and the use of low-density X-rays in the preparation of mammograms, the images have a low contrast. In addition, tumors have different sizes and shapes. Therefore, the diagnosis of lesions, especially in the early stages of the formation of work, is very difficult and tedious. The purpose of this dissertation is to investigate the localization of the breast tissue and the extraction of local features in order to detect calcification areas from calcification-free areas with higher accuracy and with the help of a computer. The results for determining the locations with and without calcification are classified with a polynomial kernel with a 100% accuracy and a sensitivity of 96% and a detection capability of 100%, which indicates the potential of the proposed method to other articles.
Keywords : Calcification Breast ultrasound imaging Feature Extraction Classification