Processing Thermographic Images for the Pre Diagnosis of Breast Cancer

Abstract

Breast cancer is a disease that begins when the cells of the breast begin to grow uncontrollably. It can not be predicted, because the factors that mainly affect the suffering of this pathology are genetic, they can be suffered by both women and men [1]. It is the most common cancer among women worldwide and accounts for 16% of all female cancers. In Colombia, it is estimated that around 8600 cases are reported every year and 2,660 women die due to this [2]. Temperature variations in the skin are organic indicators of several types of cancer, including carcinoma-type breast cancer. The most widely used method for the detection of breast cancer is mammography, a radiation that provides an anatomical image of the breast, with which atypical formations can be observed in the tissue. Mammography, “the use of X-rays, which are ionizing radiation, can cause problems, you can not take more than three or four X-ray tests per year, pregnant women can not have mammograms.” For this reason, other diagnostic alternatives have been sought. In this scenario is where infrared thermography appears, technique that allows to evaluate the behavior and the physiognomy of tissues using the temperature they have as a base, diseased tissues such as those affected by some type of cancer can present different temperature levels. If they are compared with those that are healthy. When taking this into account, breast cancer is a disease that can be characterized, detected and analyzed by thermographic images. In this work an algorithm was developed to detect breast cancer through the segmentation and processing of thermographic images of the breast. The digital process is based on the identification of the area of ​​interest, the segmentation of the color and the quantitative discrimination by color tonality. It is contained in a graphical interface to facilitate its use.