Ductal carcinoma in situ detection in breast thermography by extreme learning machine and combination of statistical measure and fractal dimension

Ductal carcinoma in situ (DCIS) is a severe breast disease. It generates little symptom and may be neglected in prodromal stage. In this study, we developed a novel DCIS detection method based on breast thermography, which can provide earlier alert than other exams. We created a 40 breast-thermogram dataset. We used six statistical measures, and we used fractal dimension to describe the texture measure. The extreme learning machine was used as the classifier. Our developed system yielded a sensitivity of 93.0 ± 2.6%, a specificity of 92.5 ± 2.6%, and an accuracy of 92.8 ± 1.8%. The extreme learning machine was better than support vector machine, artificial neural network, decision tree, and weighted k-nearest neighbors. Besides, our developed system was superior to six state-of-the-art approaches.