The application of the fuzzy theory to the clinical diagnosis of the optokinetic nystagmus

Abstract

Optokinetic nystagmus (OKN) test is popularly performed to diagnose the disequilibrium-disorders on clinical examination. Only the experts of this diseases can, however, diagnose the test result exactly, which is expressed by patterns of nystagmus requiring the pattern recognition. We constructed the computer assisted instruction system of pattern recognition using the numerical information consisting of 6 variables from the 29 normal and 22 abnormal patterns of the OKN diagnosed by one expert. The theoretical bases were Fuzzy theory especially fuzzy reasoning; if-then rule and max-min method. The high consistent rate (96%) of diagnosis was obtained between the result of the expert and of our system with this data. A lower rate (83.7%) of consistency was calculated with another 251 numerical data of OKN using this rule of fuzzy reasoning. The degree of normality and abnormality were, however, retained in this rule of reasoning. It was concluded that the fuzzy theory was useful to construct the computer assisted instruction system but the subsequent analysis was required to obtain the higher rate of consistency.