Applying a computational intelligence method to predict the rehabilitation treatment for females with lateral patellar displacement

Abstract

A lateral positioned patella has long been regarded as a major contributing factor in the development of patella femoral pain (PFP). Despite extensive research, there is still little consensus as to the most effective treatment strategy for the management of patients with lateral patellar displacement (LPD). Computational intelligence methods are proving useful aids to physicians and other medical staff, improving objectivity when making diagnostic and treatment decisions and reducing the time to make decisions. This study describes an adaptive network-based fuzzy inference system (ANFIS) used to build a model for the indirect prediction of rehabilitation treatment outcomes for females with LPD from only demographic and clinical characteristics. The prediction abilities offered using two ANFIS models are presented using data from 48 female patients referred to rehabilitation clinics of Isfahan Ayatollah Kashani and Al Zahra hospitals, Iran. The results indicate that the ANFIS model has strong potential to improve indirect prediction of rehabilitation treatment for females with LPD with a high degree of accuracy and robustness.