Narrow Results

We report the results of a comparative study of Fourier domain analysis (FDA) and texture analysis (TA) of optical coherence tomography (OCT) images of resected human breast tissues for binary classification between normal–abnormal classes and benign–malignant classes. With the incorporation of Fisher linear discriminant analysis (FLDA) in TA for feature extraction, the TA-based algorithm provided improved diagnostic performance as compared to the FDA-based algorithm in discriminating OCT images corresponding to breast tissues with three different pathologies. The specificity and sensitivity values obtained for normal–abnormal classification were both 100%, whereas they were 90% and 85%, respectively for benign–malignant classification.

Breast cancer is one of the main causes of women’s death. Thermal breast imaging is one the non-invasive method for cancer at early stage diagnosis. In contrast to mammography this method is cheap and painless and it can be used during pregnancy while ionized beams are not used. Specialists are seeking new ways to diagnose the cancer in early stages. Segmentation of the breast tissue is one of the most indispensable stages in most of the cancer diagnosis methods. By the advancement of infrared precise cameras, new and fast computers and nouvelle image processing approaches, it is feasible to use thermal imaging for diagnosis of breast cancer at early stages. Since the breast form is different in individuals, image segmentation is a hard task and semi-automatic or manual methods are usual in investigations. In this research the image data base of DMR-IR has been utilized and a now automatic approach has been proposed which does not need learning. Data were included 159 gray images used by dynamic protocol (132 healthy and 27 patients). In this study, by combination of different image processing methods, the segmentation of thermal images of the breast tissues have been completed automatically and results show the proper performance of recommended method.