Analysis of concomitant mechanisms in stroke pathogenesis using fuzzy clustering techniques

This paper looks at the potential interactions between blood, heart, and vessel conditions in the stroke patient. The present study tries to define the degree of severity of each of three groups of diagnostic categories, blood, heart, and vascular while allowing more than one to be present at once. Each condition was given a fuzzy value in the range [0,1] according to expert opinion of the severity of the condition. Each patient was mapped as a set of n=3 elements in the fuzzy hypercube of n=3 dimension. Patient clusters in the cube may represent the synergistic interactions of n variables and as such be worthy of study or used in inter-cube studies to understand other phenomena related to stroke such as outcome, medication efficacy, genetic profile, to name a few. Two experts evaluated diagnostic tests for 163 stroke patients. This data was clustered using the fuzzy c-means algorithm to search for structure and relationships in the data. Preliminary results show that there is a correlation when both the blood and vessel tests indicate a moderately severe condition in many patients.