Nasal septal perforations (NSPs) are relatively common. They can be problematic for both patients and head and neck reconstructive surgeons who attempt to repair them. Often, this repair is made using an interpositional graft sandwiched between bilateral mucoperichondrial advancement flaps. The ideal graft is nasal septal cartilage. However, many patients with NSP lack sufficient septal cartilage to harvest. Harvesting other sources of autologous cartilage grafts, such as auricular cartilage, adds morbidity to the surgical case and results in a graft that lacks the ideal qualities required to repair the nasal septum. Tissue engineering has allowed for new reconstructive protocols to be developed. Currently, the authors are unaware of any new literature that looks to improve repair of NSP using custom tissue-engineered cartilage grafts. The first step of this process involves developing a protocol to print the graft from a patient's pre-operative CT.
In this study, CT scans were converted into STereoLithography (STL) file format. The subsequent STL files were transformed into 3D printable G-Code using the Slic3r software. This allowed us to customize the parameters of our print and we were able to choose a layer thickness of 0.1mm. A desktop 3D bioprinter (BioBot 1) was then used to construct the scaffold.
This method resulted in the production of a PCL scaffold that precisely matched the patient’s nasal septal defect, in both size and shape. This serves as the first step in our goal to create patient-specific tissue engineered nasal septal cartilage grafts for NSP repair.
Bioprinting of tissue has its applications throughout medicine. Recent advances in medical imaging allows the generation of 3-dimensional models that can then be 3D printed. However, the conventional method of converting medical images to 3D printable G-Code instructions has several limitations, namely significant processing time for large, high resolution images, and the loss of microstructural surface information from surface resolution and subsequent reslicing. We have overcome these issues by creating a JAVA program that skips the intermediate triangularization and reslicing steps and directly converts binary dicom images into G-Code.
In this study, we tested the two methods of G-Code generation on the application of synthetic bone graft scaffold generation. We imaged human cadaveric proximal femurs at an isotropic resolution of 0.03mm using a high resolution peripheral quantitative computed tomography (HR-pQCT) scanner. These images, of the Digital Imaging and Communications in Medicine (DICOM) format, were then processed through two methods. In each method, slices and regions of print were selected, filtered to generate a smoothed image, and thresholded. In the conventional method, these processed images are converted to the STereoLithography (STL) format and then resliced to generate G-Code. In the new, direct method, these processed images are run through our JAVA program and directly converted to G-Code. File size, processing time, and print time were measured for each.
We found that this new method produced a significant reduction in G-Code file size as well as processing time (92.23% reduction). This allows for more rapid 3D printing from medical images.