Journal of Advances in Mathematics and Computer Science

During an investigation at the Kinkole Reference Hospital laboratory, where data was collected, it was found that the viral hepatitis screening process is particularly lengthy. This study aims to develop soft computing based on fuzzy logic and machine learning techniques to reduce the time taken to screen for viral hepatitis using medical detection strips. Based on a study conducted at Kinkole Reference Hospital, the slowness of traditional tests (45 minutes per patient for types A, B, and C) with strips was observed. The purpose is to propose an automatic system based on Python, combining several fuzzy classifiers (FCM, fuzzy SVM, ANFIS, NEFCON) to accelerate diagnosis while maintaining high reliability. The approach is based on a mathematical formalisation of the problem as a supervised classification task, optimised by minimising the squared error. In particular, the use of fuzzy SVM according to the ANYTA mukawa lukenzu approach enabled the integration of the degrees of belonging of patients to pathological classes, improving the robustness of the model in the face of uncertainty and imprecision of medical data. This software provides a more nuanced and progressive view of diagnosis. It allows the expression of the probability of belonging to a pathological state, which is particularly suited to the real clinical context where the data are rarely clear-cut. One of the major strengths of this article is precisely the rigorous presentation of the mathematical foundations before the computational implementation. This reinforces the scientific credibility of our work. In practice, this software was carried out with the implemented source codes of FCM segmentation of 63 observations, knowing their following inputs: Age, Sex, GOT and GPT, where GOT and GPT are the medical laboratory examinations with the aim of predicting the classes of viral hepatitis and their degrees of belonging. According to the software results, Hepatitis A = class 1, Hepatitis B = class 2 and Hepatitis C = class 2. These approaches pave the way for the implementation of a medical decision support system, capable of offering pre-diagnoses in real time, thus reducing the screening time from 45 minutes to a few seconds.