Patient-Specific Sensitivity of Osteotomy Interface Modelling in Mandibular Reconstruction

This study systematically investigates the influence of different interface modelling assumptions on biomechanical predictions using patient-specific models and evaluates their agreement with postoperative configurations. Five mandibular reconstruction cases using vascularised osseous free flaps were analysed. For each case, preoperative VSP-based finite element models incorporating four interface conditions, rigid osseous contact, soft-tissue compliance, initial osteotomy gap, and heterogeneous bone properties, were compared against corresponding postoperative models reconstructed from CT data acquired approximately two weeks after surgery. Reconstruction plate stress distributions under standardised masticatory loading conditions were assessed. The results demonstrate that interface modelling significantly affects both the magnitude and spatial distribution of predicted stresses. The initial osteotomy gap condition consistently produced the highest stresses (approximately 30% higher), representing a conservative upper bound, whereas the soft-tissue compliance model showed the closest agreement with postoperative outcomes. In contrast, rigid contact tended to overestimate local stress concentrations, and incorporating heterogeneous bone properties resulted in only minor differences. Patient-specific anatomical characteristics, particularly the proportion of graft relative to the native mandible, further influenced stress distribution patterns. These findings highlight the importance of physiologically realistic interface representation in improving the predictive accuracy of VSP-FEA and suggest that simple anatomical descriptors may serve as practical indicators for stratifying mechanical risk, thereby supporting more reliable and clinically relevant surgical planning.