Method and software product allowing for direct conversion of any three dimensional image (e.g., CT, MRI, ultrasound, PET, etc.) into 3D printable instructions (i.e., G-Code).
Medical Digital Imaging and Communications in Medicine (DICOM) files generated by CT or MRI scanners can be utilized to generate the instructions for 3D printers. However, the conventional method has several limitations such as significant user interaction, image manipulation including manual segmentation of the region of interest, high processing time, computer memory restrictions, and loss of structural information due to the required intermediate STL conversion before generating G-code. These limitations are compounded by common printability issues—such as large empty-space regions or background voxels—causing printers to generate incorrect toolpaths or begin printing in midair.
Dr. Rajapakse at the University of Pennsylvania has developed a streamlined end-to-end pipeline that allows users to rapidly convert medical images into ready-to-print G-code models with minimal interaction. The pipeline guides the user from DICOM import through semi-automatic segmentation, artifact removal, and printability checks, and then directly generates G-code without going through STL conversion. This substantially reduces the time to obtain a 3D print and provides greater customizability—from changing print speeds to allowing for different extrusion amounts for either increased porosity or better adhesion to varying print paths. The system accounts for parameters such as printer resolution, speed of the extruder, and method of printing, from linear to any inputted angle of rotation between each layer. These updates also incorporate volumetric-density preservation and automated handling of background or empty-space regions, ensuring reliable prints and improved ease of use for both clinical and research applications.