Fuzzy Filtering for an Intelligent Interpretation of Medical Data

This study is concerned with an intelligent interpretation of medical data in the sense that involved complexities and uncertainties (arising in understanding the data behavior) are properly (i.e. mathematically) handled. The uncertainties in data interpretation may arise due to e.g. different behavior of individuals due to the different body conditions. We use a fuzzy model to filter out the uncertainties. The fuzzy model provides an interpretation of the data without the interpretation results being affected by the uncertainties. Such a fuzzy model (that could filter out the uncertainties) is identified using a robust identification algorithm. It was demonstrated through a real-world case study that the proposed approach of fuzzy filtering is suitable for dealing with the uncertainties involved in data interpretation. Fuzzy models, due to their capability of approximating nonlinear input-output mappings, could be exploited for the filtering of uncertainties. The efficient design of the fuzzy filter is the bottleneck of the approach.