Aphasia diagnosis is a challenging medical diagnostic task due to the linguistic uncertainty and vagueness, large number of measurements with imprecision, inconsistencies in the definition of Aphasic syndromes, natural diversity and subjectivity in test objects as well as in options of experts who diagnose the disease. In this paper we present a new self-organized multi agent system that diagnoses different types of Aphasia based on fuzzy probabilities. In the proposed multi agent system, the characteristic of self organization is employed as both a decision making feature selection paradigm as well as a mechanism to estimate the probability mass functions of Aphasia factors. The estimated probability mass functions are involved in fuzzy probability calculation of different types of Aphasia. The performance and robustness of the proposed method is compared with several earlier approaches. While the proposed method requires more of the available test parameters, the comparison clearly shows the superiority of the proposed method in terms of accuracy as well as robustness.

Keywords