Neuro-fuzzy Approach for Prediction of Neurological Disorders: A Systematic Review

Neurological disorders are highly prevalent worldwide, and contributes greatly to global disease problems. This paper aimed to; presents a systematic review of the neuro-fuzzy system (NFS) approach to diagnose neurological disorders, and to identify the most effective NFS approach to guide clinical practice. A systematic search was conducted on some selected scientific databases as appropriate databases for the review. The preferred reporting items for systematic reviews and meta-analyses (PRISMA) method was used as the basic method for conducting this review. The studies selected include those in which the NFS-related approaches were developed to diagnose neurological disorders, using datasets from patients. A total of 313 papers were examined for the study, 36 papers were included according to the inclusion criteria. The results showed that among all the designed NFS models, Gaussian was the most frequently used membership functions followed by triangular functions in the selected studies, and the inference method frequently used was the Sugeno method followed by the Mamdani method. The review reveals that NFS models used are effective in diagnosing neurological disorders. However, despite the effectiveness of this approach, its usage did not cover all the neurological domains of disease diagnosis. © 2021, The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd.