Medical diagnosis using interval type-2 fuzzy similarity measures.

Concerning decision support in specific disease areas, making faster and more accurate medical diagnoses has become more challenging. Due to indeterminate, inconsistent, and huge patient data, judging a specific disease may be challenging. Similarity is an effective method to enhance healthcare systems by simplifying medical diagnostic generation compared with illness profiles. This paper proposes a new medical diagnosis algorithm that uses interval type-2 fuzzy similarity measures (IT-2FSMs). This algorithm calculates the degree of similarity between varieties of components of patient data and establishes methods of clustering patients based on close distances between some of their features. The proposed approach will be evaluated on UCI Machine Learning Repository medical datasets. Comparative studies will be done between different IT-2 FSMs (Cherif, Jaccard, Bustince, Mitchell, Gorzalczany, Zeng and Li) to demonstrate the capability of IT-2FSMs to make quick medical diagnoses, even with different noise levels. The proposed IT-2FSMs are utilized as a clustering method and compared against existing clustering algorithms such as type-2 fuzzy c-means, cluster forest, bagged clustering, evidence accumulation, and random projection. The IT-2FSMs demonstrate a pertinent classification accuracy comparable to the other algorithms, as assessed by the clustering quality parameter R. [ABSTRACT FROM AUTHOR]