An artificial intelligence based referral application to optimize orthodontic referrals in a public oral healthcare system

Abstract

An artificial intelligence expert system called the Computational Formulation of Orthodontic referral Decisions (CFOD) was created in order to provide decision-making for orthodontic diagnosis and referral in the Emirates public oral healthcare system. Purpose: The study purposes were to validate and implement the CFOD system, and to quantify GP and pediatric dentist referral patterns to the orthodontist specialist before and after implementation of the CFOD system. Materials & Methods: The CFOD system was created using eight mandatory malocclusion variables and a technology and programming stack including PHP, Codelgniter, MySQL and WampServer. The CFOD system was implemented and clinical dentists were trained following validation by 15 experienced orthodontist experts. Pre- and post CFOD system orthodontic referral patterns were analyzed. Results: Seven of eight mandatory malocclusion variables (except crowding) were directly correlated (r = 1) with orthodontist expert evaluation which validated the CFOD system. A comparison of referral patterns of orthodontic patients before and after implementation and training of the CFOD system demonstrated a 10-fold increase for GP referrals and 2-fold for pediatric dentist referrals. Conclusions: An artificial intelligence expert system for orthodontic patient referral can improve the efficiency of a large public oral healthcare system such as found in the United Arab Emirates.