Both US and German scientists have come together now to find improved methods of 3D printing such models and guides, publishing their findings in ‘From Improved Diagnostics to Presurgical Planning: High-Resolution Functionally Graded Multimaterial 3D Printing of Biomedical Tomographic Data Sets.’ While the authors point out how useful anatomical structures are in the medical field, they also explain why 3D printing medical models is currently rife with challenges:
“Typically, features of interest in source biomedical data sets (e.g., crosssectional raster-based biomedical images or volumes from MRI or CT DICOM files) are first isolated from the surrounding tissues, either through a time-consuming manual segmentation approach or through the use of specific intensity thresholds that best capture the feature(s) of interest,” state the authors. “The result of this filtering step is used to generate a single isosurface that is then translated into a printable format [i.e. stereolithography (.stl) file, the current standard for 3D printing workflows], which describes surface geometries through mesh representations of face vertices and normals.”
3D printing of surgical models can be flawed due to reasons like thresholding, or error in digital structure that prevent files from printing and then require users to make additional changes later. The researchers see multimaterial 3D printers as a solution due to their ability to use more than one .stl file:
“This approach greatly expands the number of different anatomical structures that can be represented by different colors or materials (e.g., through the use of multiple thresholds) and, as such, can improve the interpretive value of 3D-printed anatomical models,” state the researchers. “However, the printing of multiple nested .stl files is labor intensive, computationally expensive, and ultimately nonscalable.”
As a result, their study is centered around showing how functional bitmap-based workflows are helpful in fabricating biomedical data sets, permitting rapid creation of anatomical models without thresholding. The scientists also ‘bypassed’ more typical 3D printing workflows due to their potential for error, and worked with both MRI and CT data.
[Source / Images: From Improved Diagnostics to Presurgical Planning: High-Resolution Functionally Graded Multimaterial 3D Printing of Biomedical Tomographic Data Sets]“Despite the advantages of this bitmap-based workflow compared with its traditional .stl-based counterpart, there are still some limitations to the technique,” concluded the researchers. “For example, although our 3D-printed models can faithfully replicate the details of their native CT or MRI source files, their ability to successfully depict anatomical features of interest is directly limited by the inherent resolution of each imaging modality.
“Inaccuracies can also be introduced by image reconstruction methods designed to optimize human image viewing rather than 3D printing, and traditional clinical tomography file compression and archival protocols will, in turn, need to be modified as biomedical 3D printing becomes more mainstream. Data set-specific issues also arise with regard to signal intensity and range.”