i-DetectBC: Intelligence detection of breast cancer

Detecting breast cancer lesions at an early stage can help to improve the patients' survival rates. Digital mammograms can be used to detect breast cancer lesions. However, mammographic images suffer from low image quality due to the low exposure factors used. This paper proposes an interactive way of enhancing mammographic images while improving the detection of breast cancer lesions. The Intelligence Detection of Breast Cancer (i-DetectBC) allows the radiologist or clinicians to enhance the original mammographic images by using appropriate algorithms automatically. The i-DetectBC consists of two digital image processing techniques: Fuzzy Anisotropic Diffusion Histogram Equalization Contrast Adaptive Limited (FADHECAL) enhancement and Multilevel Otsu Thresholding segmentation technique. The interface of i-DetectBC can be considered user-friendly with low computational methods to provide fast results, especially when identifying breast cancer types such as benign or malignant. The i-DetectBC has been performed on 322 mammographic images, which were retrieved from the MIAS database. The efficiency of the i-DetectBC is 95.7%, and the error rate is 4.3%. In summary, this i-DetectBC can be helpful in the detection and categorization of breast cancer lesions. (paper)