An Adaptive Neuro-Fuzzy Inference System (ANFIS) Model for Prediction of Optimal Dose In Methadone Maintenance Therapy

Abstract

Methadone Maintenance Therapy is a therapy of methadone drug substitution to manage the opioid dependence in drug addictions. Prescribing the optimal dose of methadone is crucial and complex. It involves avoiding the opioid withdrawal consequences, suppressing cravings, and preventing relapse to illicit opioids use. Hence, this study provides prediction for methadone optimal dose by applying the adaptive neuro-fuzzy inference system (ANFIS) technique. ANFIS has the advantage of knowledge learning capabilities from both clinical data and the expert knowledge (i.e. fuzzy rules), and it is well suited for managing complex problems. The ANFIS model was developed by using MATLAB (i.e. Fuzzy Logic Toolbox). The results obtained show close agreement between the predicted and actual optimal doses as the obtained coefficient value are close to unity in both in both training and testing datasets. This indicates the ANFIS model is able to deal with real clinical data and it is viable to be used for predicting the optimal methadone doses in MMT patients.