Accurate and rapid diagnosis of brain tumors remains a significant challenge in modern clinical settings. While histological analysis remains the gold standard for diagnosis, issues such as limited access to anatomical data or the need for additional validation highlight the need for alternative computational approaches. This research explores the combined use of in vitro ¹H magnetic resonance spectroscopy (MRS) and soft computing (SC) methods to facilitate non-invasive tumor detection and classification. Soft computing methods such as fuzzy logic, neural networks, and evolutionary algorithms and probabilistic models – provide adaptive, cost-effective, and resilient solutions by managing imprecision and uncertainty in a way that mimics human reasoning. The focus of this study is the Adaptive Neuron-Adaptive Neuron- Fuzzy Inference System (ANFIS) is used to model complex nonlinear diagnostic relationships and aid decision making under uncertainty. Furthermore, multiple criteria decision making (MCDM) approaches - in particular weighted sum model (WSM) - are used. fine-tune diagnostic decisions based on various clinical and computational parameters. The integration of SC and MCDM provides a robust decision-support framework capable of managing complex and ambiguous clinical data. This method not only complements histological assessments, but also improves diagnostic reliability in situations where conventional histopathology is limited or unavailable.
A Thu, study studied this question.