From computed tomography to 3D printed heterogeneous scaffolds: A novel approach to simulating bone characteristics using octree adaptive subdivision

Bone microstructures or scaffolds mimicking natural bone characteristics find widespread applications in bone tissue engineering (BTE). To accurately replicate a personalized bone structure, we developed a heterogeneous scaffold by combining medical image data with the octree adaptive subdivision technique. The transformations between Hounsfield units (HU)-strength and strength-lattice features were implemented for mapping bone information to the scaffold. Initially, HU values from computed tomography (CT) images were smoothly converted into corresponding strengths through a rational Bezier curve. Then, these strengths were further translated into the features of lattice structures, specifically the strut diameter and subdivision level. The novel concept of adaptive octree subdivision, executed on the lattices, allows for controlling bone porosity and mechanical property even at local levels. As a result, a personalized heterogeneous scaffold accurately simulating natural bone structure was designed and manufactured for bone grafting application on a 3D-printed alveolar model. The scaffold has an outer cortical bone layer capable of withstanding pressures of up to 208.2 MPa, while the inner cancellous bone exhibits a high porosity ranging from 77.6 % to 93.6 %. The research provides an elegant yet effective approach to the generation of bone microstructures, opening up practical BTE applications.