The Diagnosis of Diabetic Nephropathy using Neuro- Fuzzy Expert System

Objectives: To develop an improved expert system for the diagnosis of nephropathy. Methods/Statistical Analysis: To achieve this objective, data on the nephropathy is taken by specialist doctors in this domain and adaptive neuro-fuzzy technique is applied on it. Gaussian membership functions are attempted in the study and MATLAB is used to implement the expert system. Findings: This system succeeds up to 96.25% of the cases. The sensitivity, specificity and precision obtained from this system are 97.5%, 95% and 95.12%. These parameters are found out by comparing the output achieved from this system with the judgments made by experts in this area. Application/Improvements: This expert system can be applied in the situations where the patient is unable to get medical assistance from doctor due to certain problems like low ratio of doctor to patient, unavailability of doctors in undeveloped areas etc.