The purpose of our study is to expedite cancer diagnosis through the development of software for rapid detection of hereditary breast cancer (BC) with negative BRCA1/2 on MATLAB, utilizing a fuzzy logic system with several variants of genes associated with BC. This system serves as a clinical decision-support tool, assisting in early classification and interpretation of genetic variants by combining clinical and genetic data. Clinical data were obtained from Erciyes University Faculty of Medicine Department of Medical Genetics and Uludağ University Faculty of Medicine Department of Medical Genetics. 488 individuals were studied. Only 90 of them were relevant to our investigation since their BRCA1/2 genes did not exhibit notable genetic mutations. We examined 16 distinct breast cancer risk factors and focused on mutations related to 18 hereditary BC genes. The collected data were integrated into the developed system, and various membership functions were given varying degrees of possibility, ranging from 0 to 1, depending on their participation in input clusters. After the system was trained on 90 cases and validated on six independent patients, its accuracy was assessed, yielding reliable results. Following the training phase, outcomes revealed the presence of two pathogenic variants at 0.92 (92%), two benign variants at 0.25 (25%), and two variants of unknown significance at 0.5 (50%). Given the high incidence of breast cancer, early prediction is paramount. Despite the emergence of fuzzy logic systems in medical applications, limited research akin to our study exists. The establishment of this artificial intelligence software holds promise for advancing the early detection of BC in future clinical applications.
Aliyeva et al. (Sun,) studied this question.