Breast cancer is one from various diseases that hasgot great attention in the last decades. This due to the numberof women who died because of this disease. Segmentation isalways an important step in developing a CAD system. This paperproposed an automatic segmentation method for the Region ofInterest (ROI) from breast thermograms. This method is basedon the data acquisition protocol parameter (the distance fromthe patient to the camera) and the image statistics of DMR-IRdatabase. To evaluated the results of this method, an approach forthe detection of breast abnormalities of thermograms was alsoproposed. Statistical and texture features from the segmentedROI were extracted and the SVM with its kernel functionwas used to detect the normal and abnormal breasts basedon these features. The experimental results, using the bench-mark database, DMR-IR, shown that the classification accuracyreached (100%). Also, using the measurements of the recall andthe precision, the classification results reached 100%. This meansthat the proposed segmentation method is a promising techniquefor extracting the ROI of breast thermograms
Detection of Breast Abnormalities of Thermograms based on a New Segmentation Method,. Available from: https://www.researchgate.net/publication/283423225_Detection_of_Breast_Abnormalities_of_Thermograms_based_on_a_New_Segmentation_Method [accessed Jun 2, 2016]. |
