New methodology for disability assessment: analysis of WHO-Disability Assessment Schedule II with clustering based on rules

This thesis tries to give an answer to an open question about functional disabilities (FD), constituting and application of Artificial Intelligence to medicine. In fact, there still is a lack of consensus on the concept of FD and many efforts are done at present to forward research on this field, even from the World Health Organization (WHO). On the other hand, in the context of Data Mining it is well known that some complex Knowledge Discovery (KDD) problems require combination of several techniques coming from different research areas to be properly solved. In this work a hybrid KDD technique called clustering based on rules (ClBR) has been used to analyze a database referent to the assessment of FD by means of the WHO-DASII scale, which is a new assessment scale proposed by the WHO for validating functional disability degree. After analysis and interpretation of the results, a proposal of a new taxonomy of disabilities from a real functional point of view is presented as well as its relationship with the total score of the WHO-DASII scale.