Stress-informed parametric shape adaptive patient-specific implant design in computer-assisted jaw reconstruction

Reconstructing the jaw after tumor removal or trauma increasingly relies on custom 3D-printed titanium plates. This customization applies to the implant's shape. It usually does not apply to the internal structure. The width and thickness of the plate are still determined by the designer, based on experience and clinical observation. The critical design parameters of the PSI govern its mechanical performance. Hence, if the design could analyze these parameters and adjust them accordingly, it could help avoid plate fracture. This study introduces a parametric-design approach that closes this gap.

The proposed algorithm addresses these challenges by using stress data derived from finite element analysis (FEA) to selectively reinforce only the high-stress regions of a reconstruction plate, rather than uniformly reinforcing the entire plate. The reinforcement is smoothly integrated into the plate's geometry to prevent abrupt changes in cross-section that could create stress concentrations. The approach was applied to a baseline plate at two regions of elevated stress.

Plate A reinforced only the primary zone; this reduced peak stress by 36.0% and strain by 43.8% in that zone, the largest reduction observed across both design variants, and also lowered stress and strain in the second, unmodified zone by 12.7% and 14.3%, respectively. Plate B reinforced both zones; this further reduced stress and strain at the second zone by approximately 30.6% and 33.3%, respectively, and reduced strain at the primary zone by approximately 40.6% relative to baseline. Relative to Plate A, Plate B showed a slightly higher stress at the first zone; this stress remained well below the original baseline value. This work shows that patient-specific implant geometry can be shaped directly by biomechanical feedback. This algorithm offers an efficient computational workflow, enabling implants that are not only anatomically tailored but also mechanically optimized as per given loading conditions.