Abstract Background: Chronic kidney disease (CKD) is a dynamic disease with covert nature in initial stages that often does not show any specific symptoms before advanced stages. Accurate prediction of CKD progression over time is necessary for reducing the costs and mortality rates among CKD patients. The present study proposes an adaptive neuro fuzzy inference system (ANFIS) for predicting the renal failure timeframe of CKD based on a 10-year real clinical data. Methods: This is a cohort observational study conducted on a 10-year clinical records of newly diagnosed CKD patients (n=465). The glomerular filtration rate (GFR) value was used as the marker of CKD progression and the threshold value of 15 cc/kg/min/1.73m2 defined as renal failure time point. The Takagi-Sugeno type ANFIS used a 10-year real database to predict future GFR values at three sequential time points with 6-month interval. Ten clinical and physical parameters including age, sex, weight, underlying diseases, diastolic blood pressure, creatinine, calcium, phosphorus, uric acid, and GFR were initially selected as the inputs of the predicting model. Pearson correlation coefficients test was used to determine the most significant input variables to be introduced in the ANFIS model. The fuzzy c-means clustering was used to fuzzify the variables and the membership functions were Gaussian. The predicted GFRs were compared with the real data. Results: Four variables, with significant correlation with the future GFR value, were selected as the inputs: weight(r=0.812), diastolic blood pressure(r=0.714), current GFR(t) (r=0.519), and diabetes mellitus as underlying disease (r=0.251). The comparisons of the predicted values with the real data showed that the ANFIS model could accurately estimate GFR variations in all sequential periods (normalized mean absolute error lower than 5%). Furthermore, despite increasing the prediction interval to 18-month, the ANFIS could still predict the GFR variations with high accuracy (4.88% NMAE). Conclusions: Despite the high uncertainties of human body and dynamic nature of CKD progression, the ANFIS model can predict the GFR variations at long future periods. A medical decision support system can be designed based on this model to predict the renal failure progression with acceptable accuracy through entering the basic information of a patient to adopt an appropriate treatment. Including other significant variables can improve the predicting model.
کلید واژگان :Keywords: Adaptive Neuro Fuzzy Inference System, Chronic Kidney Disease, Glomerular Filtration Rate, Renal Failure Prediction, Renal Failure Progression
ارزش ریالی : 600000 ریال
با پرداخت الکترونیک