Abstract
Three-dimensional (3D) modeling and printing comprise an important tool for orthopaedic surgeons. One area in which 3D modeling has the potential to dramatically improve our understanding of biomechanical kinematics is pathologies of the patellofemoral joint, in particular trochlear dysplasia. We describe a method for creating 3D printed models of the patellofemoral joint, including computed tomography image acquisition, image segmentation, model creation, and 3D printing. The models created can help surgeons understand and plan surgery for recurrent patellar dislocations.
Technique Video
Steps for creation of 3-dimensionally printed model of patellofemoral joint. First, a weight-bearing knee computed tomography (CT) scan is performed using the Carestream Onsight 3D Extremity system. The imaging data are then uploaded into Simpleware ScanIP (version S-2021.06), software used to convert medical imaging data into 3-dimensional (3D) models. Both automated and manual segmentation techniques are used to generate a 3D “mask” of the distal femur. Once the mask is checked for accuracy, it is optimized for printing and a patient identifier is added. The model is exported from ScanIP and then imported into PreForm, software that prepares 3D models to be printed on Formlabs printers. The desired printer material, resolution, scale, and print orientation are selected, in addition the density and location of support structures. The print is then exported to the printer, and the print is started. Once the printing process is complete, the finished print is removed from the printer and undergoes a post-processing procedure. The model can now be used to analyze the patient’s trochlear morphology.
Three-dimensional (3D) modeling is an important tool for orthopaedic surgeons in spine surgery, trauma, oncology, and arthroplasty yet has received little attention in knee surgery.1, 2, 3, 4 Standard 2-dimensional (2D) radiographs and transverse-plane tomographic images from computed tomography (CT) and magnetic resonance imaging provide limited information regarding the grossly distorted anatomy and 3D variable deformity of dysplastic patellofemoral joints in patients with recurrent patellar instability who require surgical intervention. Understanding trochlear dysplasia is important in planning surgery for patellar instability patients, yet currently, surgeons use 2D metrics including the Dejour classification and tibial tuberosity–trochlear groove (TT-TG) distance to understand the underlying complex 3D deformity.5,6 Recently, however, there has been an increase in studies that use 3D printing to analyze the patellofemoral joint.7, 8, 9, 10 We describe, in detail, our techniques for 3D printing anatomic reproductions from CT and how to use these models in planning surgery for patellar instability patients (Video 1).
Technique
Patient Selection
In patients with recurrent patellar dislocations and patellofemoral pain for whom surgical intervention is being considered, it is appropriate to make a 3D model of their patellofemoral joint. Patients with the aforementioned symptoms but with “normal” TT-TG distances are particularly suitable.
CT Scan Protocol
The Carestream Onsight 3D Extremity system (Carestream Health, Rochester, NY) for weight-bearing CT knee scanning is used (Video 1). The patient is placed in the machine in a standing position (Fig 1). The door is closed, the height of the CT machine is adjusted, and the patient places his or her arms on the support bar to remain steady in a neutral standing position for scout localizer image and scan acquisition. Afterward, the patient is placed in a standing position with the knee in 20° of flexion. The patient must stand still for approximately 1.5 minutes for each CT acquisition. Additionally, the process described in this article can be used with imaging data collected by other CT scan protocols. For example, when appropriate, we collect images with patients supine and their knees flexed at 0° and 20°, which reduces radiation and motion artifact.
Fig 1.
Weight-bearing computed tomography (CT) scan acquisition. The first step of creating a 3-dimensional (3D) model of the patellofemoral joint is imaging the patient. The Carestream Onsight 3D Extremity system is used to capture a weight-bearing CT scan of the knee. The patient is asked to stand with 1 leg in the machine, after which the door is closed and the height of the machine is adjusted. The patient places his or her arms on the support bar in a neutral standing position for scout localizer image and scan acquisition. After the acquisition of scans in the neutral position, the patient remains standing but flexes the knee to 20° of flexion. Each CT acquisition takes approximately 1 minute.
Data Export
Patient image data are exported from the medical records as Digital Imaging and Communications in Medicine (DICOM) files. To protect patient privacy, the exported imaging data should be deidentified whenever possible. Furthermore, many deidentification processes do not eliminate metadata that can be used to identify patients, so we recommend storing even deidentified image data in a secure location.
Model Segmentation
The DICOM files are imported into Simpleware ScanIP software (version S-2021.06; Synopsis, Mountain View, CA), specialized software used to convert medical imaging data into 3D models. Although we primarily use ScanIP, there exist a variety of software programs that can aid in converting 2D imaging data to 3D models.11,12 These range from free online software that can be used for prints that serve a purely esthetic purpose to software with Food and Drug Administration market clearance. Our research team uses ScanIP because of its robust image processing, measurement tools, and finite element analysis capabilities. The protocol described in this article remains roughly the same for many anatomic 3D modeling programs.
Once data are imported into ScanIP, a combination of automated and manual segmentation techniques is used to generate a 3D “mask” of the distal femur (Video 1, Fig 2).13 During this process, elements of the CT scan that are to be converted into a 3D model, namely the bony anatomy, are selected. First, a Hounsfield unit thresholding tool is used to include cortical and subchondral bone within the mask (Fig 2A). A range of Hounsfield units is selected to include the entire outline of bone while minimizing the inclusion of soft tissue.
Fig 2.
Three-dimensional (3D) print generation from computed tomography files. (A) To convert imaging data to a 3D model using Simpleware ScanIP, mask thresholding and segmentation are performed by selecting a range of Hounsfield units to be included in the model of the distal femur using the thresholding tool. The range should be optimized to include only bone, without extraneous signal in soft tissue. Suboptimal thresholding is shown, requiring further alteration of the Hounsfield unit range applied. (B) The rendered 3D reconstruction can be cropped to highlight the anatomic region of interest by altering window visibility in each direction (right, left, anterior, posterior, and so on).
Next, the mask is cropped to only include the region of interest, and any soft tissue included in the mask is removed (Fig 2B). Depending on the contrast between the soft tissue and bone, this process can result in either a nearly complete mask or a mask that requires a significant degree of manual adjustment. The contrast between the bone and soft tissue is dependent on multiple factors such as the density of the bone and CT scanner settings.
Once the initial mask has been generated, the DICOM files are analyzed visually for masking errors, including areas where bone is incorrectly missing. A manual “paint” tool can be used to make fine adjustments such as this. Special attention should be paid to the anatomy of interest to ensure that it is accurate.
Model Preparation
Once the mask is checked for accuracy, it is optimized for printing. First, the internal structure of the mask is filled. Next, the mask is converted into a surface that can be exported into a format compatible with 3D printing software. The surface is then embossed with a patient identifier (Video 1, Fig 3) and exported to an STL file.
Fig 3.
The model of the distal femur is prepared for printing and given a unique embossed label for identification after the printing process. After this preparation in ScanIP, the model is exported as an STL file for printing.
Model Printing
Although many types of 3D printers can be used, it is important to use one that can print in sufficiently high resolution to capture the anatomy of interest. As such, we use Form 3B or 3BL printers (Formlabs, Somerville, MA), which can print with submillimeter resolution; however, there are other printers on the market with adequate specifications. Although the protocol described in this article uses Formlabs printers, it is generalizable to other printers as well.
The STL file is imported into PreForm software (Formlabs), software that prepares STL files to be printed on Formlabs printers (Fig 4). The desired printer material, resolution, scale, and print orientation are selected, in addition to the density and location of support structures. We tend to choose Grey v4 resin (Formlabs) as the material and a resolution of 0.1 mm for the z-axis because these suit our esthetic and detail needs, respectively. Although the choice of material is arbitrary, it is important to choose a resolution that is sufficient to capture the details of interest. As a rule of thumb, we choose a resolution higher than the resolution of the CT data so that no morphologic information is lost during the printing process. We recommend orienting the print such that the surface of the trochlea is not in contact with any support material to ensure the fidelity of the surface features. The file is then exported to the printer and printed.
Fig 4.
Model importation for printing. The STL file of the 3-dimensional model of the distal femur is imported into PreForm, a software platform compatible with the Formlabs family of 3-dimensional printers, and oriented on the baseplate for printing.
Post-processing
Once the printing process is complete, the finished print is removed from the printer and washed in isopropyl alcohol for 20 minutes in a Form Wash system (Formlabs). After the wash, the print is allowed to dry for at least 30 minutes. Next, the print undergoes a post-cure process in which it is exposed to ultraviolet light for 30 minutes at 60°C in a Form Cure system (Formlabs) (Video 1, Fig 5). Finally, the support material is removed carefully with hand tools. The print is now ready to be analyzed (Fig 6). Pearls and pitfalls for the entire technique can be found in Table 1.
Fig 5.
Three-dimensional print curing under ultraviolet light. After completion of the 3-dimensional printing process, model washing in isopropyl alcohol, and a 30-minute drying period, the 3-dimensionally printed distal femur is placed in the Form Cure system for ultraviolet treatment for 30 minutes. After completion of curing, any support struts present are removed and the model is considered ready for analysis.
Fig 6.
Completed 3-dimensionally printed model of distal femur, which can now be used to better understand patient’s trochlear morphology.
Table 1.
Pearls and Pitfalls
| Pearls |
| One should ensure that the CT data are of sufficiently high resolution to capture the anatomy of interest. |
| Osteopenic bone can have poor contrast with soft tissue and often requires more manual adjustments with the paint tool. |
| It is better to under-crop the mask than to over-crop. |
| Printing should be done using higher resolution than the imaging data to ensure that no information is lost in the printing process. |
| The print should be oriented such that support material does not contact the surface of the trochlea or any regions of interest. |
| Pitfalls |
| Many deidentification processes leave metadata that can be used to identify patients. |
| Including too large a range of Hounsfield units will lead to the inclusion of soft tissue in the mask, which can be time-consuming to remove. |
| One must be mindful of implants when thresholding because these can mistakenly be included in the mask. |
| Failing to check the accuracy of the mask after thresholding will lead to distorted and incorrect anatomy on the print. |
CT, computed tomography.
Discussion
Trochlear dysplasia is a complex 3D deformity closely related to patellar instability.14 A 2D understanding of trochlear dysplasia using the Dejour classification, as well as measures of proximal patellar flattening such as lateral trochlear inclination, the TT-TG distance, and the Caton-Deschamps ratio, is helpful in surgical planning for patellar instability patients, but 3D reproductions of dysplastic 3D geometry and dysplastic trochlear orthogonal tracking paths can further improve our understanding of trochlear dysplasia in the care of selected patellar instability patients.5,6,15 Of particular interest is the markedly improved understanding of trochlear curvatures and obliquity possible upon study of 3D prints of dysplastic trochleae.13 Understanding where a patella enters a trochlea, particularly the laterality of entry, is useful in surgical planning for tibial tubercle transfer osteotomy or femoral derotation. We believe, however, that 3D modeling has the potential to go far beyond what we outline in this article regarding understanding trochlear dysplasia, patellar instability, patellofemoral pain, and origins of patellofemoral arthritis. The method described in this article and Video 1 provides the information necessary to create 3D models selectively to help in planning patellofemoral surgery.
Three-dimensional images and prints provide a unified, comprehensive understanding of trochlear groove obliquity and curvature in recurrent patellar instability patients, as well as a structural correlate to the clinical J sign. Using 3D models, surgeons can readily follow the medial trochlear ridge deformity of trochlear dysplasia to understand where and how the patella enters the trochlea. Visualizing the 3D tracking path of a patella in a distorted trochlea and seeing the 3D relations between the tibial tuberosity and orthogonal patellar tracking path provide important insights for both designing recurrent patellar instability surgery and understanding unusual or severe deformity in other selected patients.
Acknowledgment
The authors thank Andrew Osborne and Kimberly Conner of the Yale School of Medicine Office of Communications for their assistance with the video.
Footnotes
The authors report the following potential conflicts of interest or sources of funding: This publication was made possible by the Richard K. Gershon, M.D. Fund at Yale UniversitySchool of Medicine. D.B.F. receives personal fees from Orthofix and Ultragenyx, outside the submitted work. D.H.W. receives consulting fees from Intellijoint, Globus, and Materialise, outside the submitted work J.P.F. is President of the Patellofemoral Foundation. Full ICMJE author disclosure forms are available for this article online, as supplementary material.
Supplementary Data
Steps for creation of 3-dimensionally printed model of patellofemoral joint. First, a weight-bearing knee computed tomography (CT) scan is performed using the Carestream Onsight 3D Extremity system. The imaging data are then uploaded into Simpleware ScanIP (version S-2021.06), software used to convert medical imaging data into 3-dimensional (3D) models. Both automated and manual segmentation techniques are used to generate a 3D “mask” of the distal femur. Once the mask is checked for accuracy, it is optimized for printing and a patient identifier is added. The model is exported from ScanIP and then imported into PreForm, software that prepares 3D models to be printed on Formlabs printers. The desired printer material, resolution, scale, and print orientation are selected, in addition the density and location of support structures. The print is then exported to the printer, and the print is started. Once the printing process is complete, the finished print is removed from the printer and undergoes a post-processing procedure. The model can now be used to analyze the patient’s trochlear morphology.
References
- 1.Bagaria V., Chaudhary K. A paradigm shift in surgical planning and simulation using 3Dgraphy: Experience of first 50 surgeries done using 3D-printed biomodels. Injury. 2017;48:2501–2508. doi: 10.1016/j.injury.2017.08.058. [DOI] [PubMed] [Google Scholar]
- 2.Chi-Kay L., King-him C., Kin-bong L., Wilson L. Computer-assisted planning and three-dimensional-printed patient-specific instrumental guide for corrective osteotomy in post-traumatic femur deformity: A case report and literature review. J Orthop Trauma Rehabil. 2018;24:12–17. [Google Scholar]
- 3.Lin C.Y., Wirtz T., LaMarca F., Hollister S.J. Structural and mechanical evaluations of a topology optimized titanium interbody fusion cage fabricated by selective laser melting process. J Biomed Mater Res A. 2007;83:272–279. doi: 10.1002/jbm.a.31231. [DOI] [PubMed] [Google Scholar]
- 4.Koeck F.X., Beckmann J., Luring C., Rath B., Grifka J., Basad E. Evaluation of implant position and knee alignment after patient-specific unicompartmental knee arthroplasty. Knee. 2011;18:294–299. doi: 10.1016/j.knee.2010.06.008. [DOI] [PubMed] [Google Scholar]
- 5.Kazley J.M., Banerjee S. Classifications in brief: The Dejour classification of trochlear dysplasia. Clin Orthop Relat Res. 2019;477:2380–2386. doi: 10.1097/CORR.0000000000000886. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Dejour H., Walch G., Nove-Josserand L., Guier C. Factors of patellar instability: An anatomic radiographic study. Knee Surg Sports Traumatol Arthrosc. 1994;2:19–26. doi: 10.1007/BF01552649. [DOI] [PubMed] [Google Scholar]
- 7.Fritz B., Fucentese S.F., Zimmermann S.M., Tscholl P.M., Sutter R., Pfirrmann C.W.A. 3D-printed anatomic models of the knee for evaluation of patellofemoral dysplasia in comparison to standard radiographs and computed tomography. Eur J Radiol. 2020;127 doi: 10.1016/j.ejrad.2020.109011. [DOI] [PubMed] [Google Scholar]
- 8.Pennock A.T., Chang A., Doan J., Bomar J.D., Edmonds E.W. 3D knee trochlear morphology assessment by magnetic resonance imaging in patients with normal and dysplastic trochleae. J Pediatr Orthop. 2020;40:114–119. doi: 10.1097/BPO.0000000000001188. [DOI] [PubMed] [Google Scholar]
- 9.Liao T., Jergas H., Tibrewala R., et al. Longitudinal analysis of the contribution of 3D patella and trochlear bone shape on patellofemoral joint osteoarthritic features. J Orthop Res. 2020;39:506–515. doi: 10.1002/jor.24836. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Fuchs A., Feucht M.J., Dickschas J., et al. Interobserver reliability is higher for assessments with 3D software-generated models than with conventional MRI images in the classification of trochlear dysplasia. Knee Surg Sports Traumatol Arthrosc. 2022;30:1654–1660. doi: 10.1007/s00167-021-06697-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Bugeja J.M., Xia Y., Chandra S.S. Automated 3D analysis of clinical magnetic resonance images demonstrates significant reductions in cam morphology following arthroscopic intervention in contrast to physiotherapy. Arthrosc Sports Med Rehabil. 2022;4:e1353–e1362. doi: 10.1016/j.asmr.2022.04.020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Oishi T., Kobayashi N., Yukizawa Y., Takagawa S., Honda H., Inaba Y. Radiographs are comparable with 3-dimensional computed tomography-based models as a modality for the preoperative planning of the arthroscopic lateral acromioplasty: A retrospective comparative study. Arthrosc Sports Med Rehabil. 2022;4:e1799–e1806. doi: 10.1016/j.asmr.2022.07.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Yu KE., Cooperman DR., Schneble CA., et al. Reconceptualization of trochlear dysplasia in patients with recurrent patellar dislocation using 3-dimensional models. Orthop J Sports Med. 2022 doi: 10.1177/23259671221138257. 10:23259671221138257. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Levy B.J., Tanaka M.J., Fulkerson J.P. Current concepts regarding patellofemoral trochlear dysplasia. Am J Sports Med. 2020;49:1642–1650. doi: 10.1177/0363546520958423. [DOI] [PubMed] [Google Scholar]
- 15.van Duijvenbode D., Stavenuiter M., Burger B., van Dijke C., Spermon J., Hoozemans M. The reliability of four widely used patellar height ratios. Int Orthop. 2016;40:493–497. doi: 10.1007/s00264-015-2908-2. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Steps for creation of 3-dimensionally printed model of patellofemoral joint. First, a weight-bearing knee computed tomography (CT) scan is performed using the Carestream Onsight 3D Extremity system. The imaging data are then uploaded into Simpleware ScanIP (version S-2021.06), software used to convert medical imaging data into 3-dimensional (3D) models. Both automated and manual segmentation techniques are used to generate a 3D “mask” of the distal femur. Once the mask is checked for accuracy, it is optimized for printing and a patient identifier is added. The model is exported from ScanIP and then imported into PreForm, software that prepares 3D models to be printed on Formlabs printers. The desired printer material, resolution, scale, and print orientation are selected, in addition the density and location of support structures. The print is then exported to the printer, and the print is started. Once the printing process is complete, the finished print is removed from the printer and undergoes a post-processing procedure. The model can now be used to analyze the patient’s trochlear morphology.
Steps for creation of 3-dimensionally printed model of patellofemoral joint. First, a weight-bearing knee computed tomography (CT) scan is performed using the Carestream Onsight 3D Extremity system. The imaging data are then uploaded into Simpleware ScanIP (version S-2021.06), software used to convert medical imaging data into 3-dimensional (3D) models. Both automated and manual segmentation techniques are used to generate a 3D “mask” of the distal femur. Once the mask is checked for accuracy, it is optimized for printing and a patient identifier is added. The model is exported from ScanIP and then imported into PreForm, software that prepares 3D models to be printed on Formlabs printers. The desired printer material, resolution, scale, and print orientation are selected, in addition the density and location of support structures. The print is then exported to the printer, and the print is started. Once the printing process is complete, the finished print is removed from the printer and undergoes a post-processing procedure. The model can now be used to analyze the patient’s trochlear morphology.