An evolutionary-fuzzy approach for supporting diagnosis and monitoring of Multiple Sclerosis

The diagnosis and monitoring of Multiple Sclerosis (MS) are very thorny tasks due to extremely variable and often quite subtle symptoms. The use of MR images as MS marker requires the expert´s knowledge and intervention to classify MS lesions. In this respect, the paper proposes an evolutionary-fuzzy approach aimed at supporting the classification of lesions in the diagnosis and monitoring of MS. Such an approach consists in: i) the formalization of the expert´s medical knowledge in terms of linguistic variables, linguistic values and fuzzy rules; ii) the implementation of a fuzzy inference technique to identify MS lesions and an evolutionary-fuzzy algorithm to tune the shapes of the membership functions for each linguistic variable involved in the rules. An experimental evaluation has been performed on 120 patients affected by MS.