M-Health System Framework for Diagnosing Inflammatory Breast Cancer with Fuzzy Logic

Inflammatory breast cancer (IBC) is an aggressive and fatal breast cancer. The American Joint Committee on Cancer defines the clinical symptoms of IBC as erythema, edema, peau d'orange that is over at least a third of the breast with a duration of the first symptom to the diagnosis of fewer than six months, and histopathologic diagnosis. Unfortunately, these signs are not present in many cases and make it challenging to diagnose IBC. In this work, we proposed a pioneering framework of an M-Health system for early diagnosis of IBC to improve survival rates and presented the supporting preliminary experimental results. Particular emphasis is given to the system framework and bilateral mammography images. Since IBC is a rare type of cancer, there is currently no public-domain mammography image dataset. Our work evaluated the system performance using the bilateral mammography images of six IBC and eight non-IBC breast cancer cases provided by the National Cancer Institute (NCI) - Cairo, Egypt. The proposed model extracts features and combines them in the feature bank to send to the Fuzzy logic Type 1 classifier for diagnosis. The system achieved promising performance with an accuracy of 92.3%, sensitivity of 83.3%, and specificity of 100%.