FROM THE JOURNAL OF THORACIC AND CARDIOVASCULAR SURGERY
Flexible 3-D models of 10 human patient pulmonary arteries were constructed using 3-D printing as part of a project to develop a new delivery catheter for regional lung chemotherapy.
Computed tomography and CT angiography in combination with software-driven segmentation techniques were used for generation and adjustment of 3-D polygon mesh to form reconstructed models of the pulmonary arteries. The reconstructed models were exported as stereolithographic data sets and further processed, according to Sergei N. Kurenov of the department of thoracic surgery, Roswell Park Cancer Institute, Buffalo, N.Y., and his colleagues.
In this process, producing the anatomical pulmonary artery models required a series of steps: data acquisition from the patient CT digital data, 3-D visualization and segmentation, surface rendering and creating a 3-D polygon mesh, geometrical surface preparation – simplification, refinement, and geometry fixing, and the hollowing of an existing volume to “thicken” the walls.
Three contrast CT data sets with a 0.625-mm, 1-mm, and 2-mm slice thickness were gathered for each patient.
The scans were processed using commercial software packages. Because of the high variability of curvature and embedding in complex anatomical scenes with other vessels interference, the pulmary artery segmentation using the software tools required a clear understanding of the patient’s anatomy, which took 4-8 hours for the experienced operator, according to the researchers.
After further computer processing of the virtual reconstructed pulmonary model, it was sent to the 3-D printer, which used a rubberlike material that is elastic and semitransparent, behaving similarly to polyurethane.
After printing, a power washer using a water jet was used to remove support material and material inside the large vessels. To clean vessels smaller than 2 mm or very tortuous regions, various catheters connected to a water faucet were used.
The 10 unique models were successfully created with no print failures, although the original plan of using a 1-mm mural thickness proved too fragile, so the entire group was printed with a 1.5-mm wall. The design process took 8 hours from CT image to stereolithographic model, and printing required an overall total of 97 hours, according to the report published online in the Journal of Thoracic and Cardiovascular Surgery and appearing in the April issue [ doi:10.1016/j.jtcvs.2014.12.059 ].
The physical measurements of the model were accurate for clinical purposes, with the 95% confidence levels for the 10 models demonstrating equivalence. Anatomic measurements using this process could be useful for general pulmonary artery catheter design, according to the authors. These measurements showed sufficient similarity for a design to be created that would be effective for most patients, although this finding would have to be validated with a larger sample of patients.
“While many of the measurements could have been made with software analysis of the 3-D files, some measurements were greatly facilitated by bending the model and aligning the physical catheter. These measurements represent distance beyond which a catheter might cause damage,” they added.
“Going forward, this technology competes with virtual educational media for health care professionals, trainees, and patients. Complex anatomy can be visualized easily on a scale model at the operating table (rather than by manipulating a nonsterile pointing device on a computer). The [pulmonary arteries] we printed could be used in a relatively low-cost lifelike [video-assisted thoracoscopic] lobectomy trainer,” the authors stated, while acknowledging the current issue of cost and time. “However, it is reasonable to expect its use to become rapidly more accessible given the competitive growth of this technology for both commercial and personal applications.
Printing services were funded by an unrestricted grant from Incodema3D, which employs Dan Sammons, one of the authors of the study. The other authors had no relevant financial disclosures.
