Background: Application of artificial intelligence to predict and explore potential relationship between predictors and outcome in biologic nature has been increasingly used in many clinical scenarios. The purpose of this study was to apply and validate artificial neural network (ANN) and naive Bayes classifier (NBC), two models of artificial intelligence, in predicting the target range of plasma intact parathyroid hormone (iPTH) concentration for hemodialysis patients. Methods: The study population included 130 stable hemodialysis patients. The predictors consisted of demographic characteristics (gender, age), associated diseases (diabetes, hypertension), and blood biochemistries (hemoglobin, protein, albumin, calcium, phosphorus, alkaline phosphatase, and ferritin), calcium-phosphorus product, and transferrin saturation values. Plasma iPTH concentration measured by radioimmunometric assay was the dichotomous outcome variable, either target group (150 ng/L≤iPTH≤300 ng/L) or non-target group (iPTH < 150 ng/L or iPTH > 300ng/L) on the basis of Kidney Disease Outcomes Quality Initiative guidelines. The leave-one-out cross validation was employed to surmount the generalization problem caused by a small amount of study population. To compare the performance of the ANN and NBC models, discrimination was evaluated using the area under the receiver operating characteristic curve (AUC) and calibration was estimated using the Hosmer-Lemeshow goodness-of-fit statistic (H-statistic). Results: Pairwise comparison of each AUC showed that the ANN model significantly outperformed the NBC model (AUC=0.90±0.06 vs. 0.62±0.08, P<0.01). The H-statistic values of the ANN and NBC models were 6.88 (P=0.08) and 6.97 (P=0.07), respectively. The ANN model with a lower H-statistic and a higher P value than the NBC model was associated with a better fit. Conclusion: The ANN model could serve as a promising tool to forecast the target range of plasma iPTH concentration in hemodialysis patients.
Background: Application of artificial intelligence to predict and explore potential relationship between predictors and outcome in biologic nature has been increasingly used in many clinical scenarios. The purpose of this study was to apply and validate artificial neural network (ANN) and naive Bayes classifier (NBC), two models of artificial intelligence, in predicting the target range of plasma intact parathyroid hormone (iPTH) concentration for hemodialysis patients. Methods: The study population included 130 stable hemodialysis patients. The predictors consisted of demographic characteristics (gender, age), associated diseases (diabetes, hypertension), and blood biochemistries (hemoglobin, protein, albumin, calcium, phosphorus, alkaline phosphatase, and ferritin), calcium-phosphorus product, and transferrin saturation values. Plasma iPTH concentration measured by radioimmunometric assay was the dichotomous outcome variable, either target group (150 ng/L≤iPTH≤300 ng/L) or non-target group (iPTH < 150 ng/L or iPTH > 300ng/L) on the basis of Kidney Disease Outcomes Quality Initiative guidelines. The leave-one-out cross validation was employed to surmount the generalization problem caused by a small amount of study population. To compare the performance of the ANN and NBC models, discrimination was evaluated using the area under the receiver operating characteristic curve (AUC) and calibration was estimated using the Hosmer-Lemeshow goodness-of-fit statistic (H-statistic). Results: Pairwise comparison of each AUC showed that the ANN model significantly outperformed the NBC model (AUC=0.90±0.06 vs. 0.62±0.08, P<0.01). The H-statistic values of the ANN and NBC models were 6.88 (P=0.08) and 6.97 (P=0.07), respectively. The ANN model with a lower H-statistic and a higher P value than the NBC model was associated with a better fit. Conclusion: The ANN model could serve as a promising tool to forecast the target range of plasma iPTH concentration in hemodialysis patients.