Purpose or Case Report: Three-dimensional (3D) printing typically utilizes CT or MRI data to create patient specific models for detailed visualization and surgical planning. However, relying on a single imaging modality can limit anatomical accuracy. Here, we describe novel methods to produce 3D-printed and VR models of fused multimodal imaging of three complex pediatric craniofacial vascular lesions.
Methods & Materials: To match high-resolution vascular architecture with hemodynamic and flow-sensitive imaging data, spatial registration and fusion are performed using DICOM datasets imported into Materialise Mimics. Following registration, lesions are delineated via region-growing and manual refinement, while vascular structures are segmented using thresholding methods based on Hounsfield units and signal intensities. Morphological operations are applied to remove artifacts and smooth vessel walls. Color mapping is used to distinguish vasculature. The completed segmentation is converted to a 3D triangulated mesh (STL) using marching cubes algorithm, then refined in Blender or 3-matic to ensure printability by sealing holes, removing excess polygons, and adding connectors. The final digital model is fabricated using PolyJet or SLA printing with multicolored or transparent resins, enhancing anatomical visualization and highlighting flow channels.
Results: Fusing different imaging modalities juxtaposes vascular detail with soft tissue contrast, yielding precise 3D models of complex craniofacial anatomy for surgical planning and education. Multimodal registration using Mimics defined a high-flow skullbase dural AV fistula for targeted embolization, mapped arterial feeders in a lip vascular malformation to guide safe resection, and profiled shared venous sinuses in craniopagus twins for staged surgical separation. Across the three cases, models from image fusion elucidated vascular relationships while facilitating low morbidity, high efficiency interventions.
Conclusions: This workflow marks a significant advancement in precision medicine, pushing the limits of FDA-approved software for anatomical modeling. It is reproducible, compatible with multiple platforms, and offers a cost-effective method to convert vendor-agnostic fused imaging into accurate 3D-printed and VR models. Our multimodality fusion models transformed into tangible, educational, and clinically actionable tools, bridging the gap between visualization and surgical execution.