Why detailed modelling matters in the pre-clinical evaluation of temporomandibular joint implants

Abstract

End-stage temporomandibular joint (TMJ) disorders often necessitate joint replacement to restore bilateral mastication. Patient-specific TMJ implants typically rely on detailed in silico modelling derived from subject-specific computed tomography (CT) scans of the mandible. However, generating highly detailed finite element (FE) models is computationally expensive. Although structural simplifications of the mandible are known to influence mechanical behaviour, their quantitative impact on TMJ implant evaluation has not been systematically investigated. This study examines the influence of three modelling simplifications on stress-strain predictions in intact and implanted mandibles: (i) a detailed segmented mandible with tissue-specific material properties (Model 1), (ii) a simplified mandible comprising cortical bone only (Model 2), and (iii) a further simplified mandible excluding dental crowns (Model 3). Both narrow and standard TMJ implants were analysed under osseointegrated and non-osseointegrated conditions, resulting in fifteen FE models. Physiological clenching activities were simulated. Compared with the detailed model, simplified models showed reductions of up to 50% in maximum principal strain in bone and up to 44% in von Mises stress in TMJ implants, while preserving spatial stress-strain trends. These findings indicate that simplified models may be suitable for preliminary implant design, whereas detailed FE modelling remains essential for final pre-clinical evaluation.