Quality Assurance in Medical 3D-Printing

In the strive towards patient specific healthcare, medical 3D-printing has proven to be of great value. With the growing possibilities in applicability of these 3D-prints, quality control of the complete surgical 3D-printing workflow: (1) case selection, (2) image acquisition, (3) segmentation, (4) engineering, (5) 3D-printing and (6) preparation, has never been of more importance. The objective of this study was to provide a quality control methodology to assess the reproducibility, accuracy and stability of a medical 3D-printing workflow. With computer-aided-design (CAD) software a quality control (QC)-phantom was designed with dimensional characteristics based on commonly used surgical guides. Seven identical QC-phantoms were 3D-printed and labeled (n = 1, ..., 7). Their dimensional parameters were determined by physical measurements. Through comparison with the CAD-drawing, information was gained on the performance of the 3D-printer. The seventh phantom was scanned six times on a CT-scanner with a slice thickness of 0.4 mm. The 3D-segmented phantom was 3D-printed and its characteristics compared to the original. The other six phantoms (n = 1, ..., 6) underwent a number of sterilization cycles (3 min, 134 degrees C) corresponding with their label. The overall 3D-printer accuracy due to solely printing was found to be within 0.2 mm deviation with an average of 0.04 mm and a standard deviation of 0.06 mm. Image acquisition and segmentation induced shrinkage of the slots and holes. Thereby, a deviation of maximum 0.4 mm occurred compared to the original phantom, the average deviation was 0.06 mm with a standard deviation of 0.06 mm. Furthermore, within a cycle of six sterilizations (n = 1, ..., 6) the accuracy and durability of the materials did not deteriorate. By design of this phantom, a method to implement overall QC of a medical 3D-printing workflow was developed. Execution of this method showed that the quality of the complete medical 3D-printing workflow can be assured and that it provides insight in the influences of each of the different steps in the workflow. In this study the strongest effect on the dimensional characteristics was observed in the image acquisition step.