Advertisement

Development of a 3-D printing-based cardiac surgical simulation curriculum to teach septal myectomy

Open ArchivePublished:November 06, 2017DOI:https://doi.org/10.1016/j.jtcvs.2017.09.136

      Abstract

      Objective

      We sought to develop a 3-D printing-based simulator for teaching extended septal myectomy to trainees in cardiothoracic surgery (clinical postgraduate year 4-7). This procedure is difficult to teach because of generally unfamiliar and highly variable anatomy, limited visibility for the assistant, and significant potential complications.

      Methods

      A curriculum using multimedia didactics and 3-D print-based patient-specific surgical simulation was implemented. Six identical 3-D prints were constructed for each of 5 consecutive patients. Preoperative septal myectomy was performed on each printed heart by an attending surgeon and 5 residents. Model myectomy specimen volumes were measured according to liquid displacement. All print resections were videotaped and blindly evaluated by 3 attending surgeons. Pre- and post-test evaluations, and a survey tool were also used to evaluate the curriculum.

      Results

      Baseline myectomy resection volumes differed significantly (attending 15 cm3 vs resident 3.1 cm3; P < .05). Residents resected increasingly larger volumes of tissue over the course of the study. Initial resection volume (compared with faculty) increased by 0.82 cm3 per resection (95% confidence interval, 0.37-1.3 cm3; P < .0001). Total resection volume (compared with faculty) increased by 3.6 cm3 per resection (95% confidence interval, 2.4-4.9 cm3; P < .0001). The residents’ survey assessment of the simulator was favorable.

      Conclusions

      A patient-specific 3-D printing-based simulation module shows promise as a tool to augment and improve cardiothoracic resident training in septal myectomy. The residents were quickly able to perform resections on par with the attending. Residents rated the simulator favorably. Each resident benefited by experiencing the variable anatomy of 5 separate patient-specific models.

      Key Words

      Abbreviations and Acronyms:

      CABG (coronary artery bypass grafting), CI (confidence interval), LVOT (left ventricular outflow tract), VSD (ventricular septal defect)
      Figure thumbnail fx1
      Myectomy done by attending and resident at the beginning (left) and near the end of the study (right). Black circles show the area of resection.
      A 3-D print-based simulation curriculum was used to augment cardiothoracic resident training in septal myectomy.
      Septal myectomy is a difficult procedure to teach in most training programs because of low surgical volume, generally unfamiliar and highly variable anatomy, limited visibility for the assistant, and significant specific complications. A simulation curriculum using patient-specific 3-D prints is useful to augment resident education and training for this specific procedure.
      See Editorial Commentary page 1149.
      See Editorial page 1137.
      Surgical myectomy for treatment of left ventricular outflow tract (LVOT) obstruction in patients with hypertrophic cardiomyopathy is a conceptually simple, yet challenging 3-D procedure. Septal myectomy is difficult to teach compared with many cardiac procedures because of highly variable anatomy that is generally unfamiliar to residents, limited visibility for the assistant, significant specific complications, and low surgical volume in all but a few institutions.
      • Maron B.J.
      • Rastegar H.
      • Udelson J.E.
      • Dearani J.A.
      • Maron M.S.
      Contemporary surgical management of hypertrophic cardiomyopathy, the need for more myectomy surgeons and disease-specific centers, and the Tufts initiative.
      We previously reported on the use of patient-specific 3-D printed models for deliberate practice and surgical rehearsal (by an attending surgeon) of extended septal myectomy for hypertrophic cardiomyopathy.
      • Hermsen J.L.
      • Burke T.M.
      • Seslar S.P.
      • Owens D.S.
      • Ripley B.A.
      • Mokadam N.A.
      • et al.
      Scan, plan, print, practice, perform: development and use of a patient-specific 3-dimensional printed model in adult cardiac surgery.
      The models are printed in a proprietary hydrogel medium and are able to be manipulated in ways meaningful to a surgeon. This medium is overall stiffer and less deformable than myocardium but is able to be incised, cut, and sutured. It was apparent from the initial experience that this model might also have educational utility.
      • Hermsen J.L.
      • Burke T.M.
      • Seslar S.P.
      • Owens D.S.
      • Ripley B.A.
      • Mokadam N.A.
      • et al.
      Scan, plan, print, practice, perform: development and use of a patient-specific 3-dimensional printed model in adult cardiac surgery.
      The objective of this study was to development and use a 3-D print-based simulation curriculum to augment traditional instruction of senior cardiothoracic residents regarding extended septal myectomy. We recognized our own residents were undertrained in this procedure and hypothesized that a simulation-based curriculum using patient-specific 3-D printed models might have utility.
      Within surgical disciplines, teaching and learning of specific procedures are paramount and educational paradigms and pedagogy continue to evolve. Simulation has become increasingly used as an adjunct to operating room experience and the concept of deliberate practice is gaining traction among surgeon-educators.
      • Mokadam N.A.
      • Fann J.I.
      • Hicks G.L.
      • Nesbitt J.C.
      • Burkhart H.M.
      • Conte J.V.
      • et al.
      Experience with the cardiac surgery simulation curriculum: results of the resident and faculty survey.
      • Cook M.R.
      • Graff-Baker A.N.
      • Moren A.M.
      • Brown S.
      • Fair K.A.
      • Kiraly L.N.
      • et al.
      A disease-specific hybrid rotation increases opportunities for deliberate practice.

      Methods

       Curriculum

       Cognitive assessment

      Each resident completed a 10 item pretest (Appendix E1) before the didactic lectures. The test was readministered at the completion of the curriculum (after all simulations and cases) and scores were compared.

       Didactics

      A curriculum composed of 2 didactic lectures (1 by a cardiac surgeon experienced in myectomy and 1 by a radiologist with 3-D printing expertise), selected readings,
      • Smedira N.G.
      • Lytle B.W.
      • Lever H.M.
      • Rajeswaran J.
      • Krishnaswamy G.
      • Kaple R.K.
      • et al.
      Current effectiveness and risks of isolated septal myectomy for hypertrophic obstructive cardiomyopathy.
      • Maron B.J.
      • Ommen S.R.
      • Semsarian C.
      • Spirito P.
      • Olivotto I.
      • Maron M.S.
      Hypertrophic cardiomyopathy: present and future, with translation into contemporary cardiovascular medicine.
      • Schaff H.V.
      • Said S.
      Transaortic extended septal myectomy for hypertrophic cardiomyopathy.
      and a surgical video
      • YouTube
      Video of my heart surgery (myectomy).
      was designed and implemented. The readings and video were provided electronically and the lectures given before beginning the simulation portion of the curriculum.

       Simulations

      Figure thumbnail fx2
      Video 1Video clip showing a resident myectomy on a 3-D print. The resident begins by marking the right coronary ostium (correctly), the annulus of the right coronary sinus of Valsalva (correctly), and the membranous septum (incorrectly). The rightward extent of myectomy is correctly identified as a line drawn from the right coronary ostium down to the annulus. Resection of the initial piece of septum is shown. Video available at: https://www.jtcvs.org/article/S0022-5223(17)32407-8/fulltext.
      Figure thumbnail gr1
      Figure 1Photograph showing the setup of the simulator within an empty operating room. The 3-D print is positioned within a basin placed in the chest cavity of a mannequin. A sternal retractor and drapes help to recreate the field. Basic instruments and the graduated cylinder used to measure resection volumes are shown on the Mayo stand.
      No technical instruction, other than provided in the didactic materials, was given before resection of patient 1 models. This allowed for the resident resection of that model to serve as a baseline. All residents reviewed the patient 1 attending model alongside a representative patient 1 resident model before working with the patient 2 model. Some basic technical feedback was given during patients 3 to 5 model simulation sessions. This feedback was provided by J.L.H., who was the attending surgeon on all cases and was also one of the video graders.
      The entire study was conducted during an approximate 6-month period. The time between the didactics and first patient was approximately 3 weeks. The remainder of the patients were spread relatively evenly over the ensuing 5 months.
      Faculty scored all model resection videos using a Likert scale-based tool with clinical performance anchors (Appendix E2). Scores were analyzed using mixed effects in a similar fashion. All faculty assessment scores were averaged to provide a mean faculty assessment score for each resident, for each model. Assessment of faculty agreement was performed using the mean faculty assessment score adjusting for fixed effects at the resident and model levels. A logistic regression model was used to model faculty status according to faculty assessment mean scores.

       Survey assessment

      An anonymous survey tool was developed, on the basis of a tool used for a similar simulation study,
      • Barsness K.A.
      • Rooney D.M.
      • Davis L.M.
      Collaboration in simulation: the development and initial validation of a novel thoracoscopic neonatal simulator.
      and administered after completion of the curriculum. The tool asked residents to assess several domains of the simulator including physical attributes and realism, ability to perform key tasks, readiness for educational use, value as a teaching tool, and worthiness related to effort and future practice.
      Institutional review board approval was waived for this study. All residents assented to full participation in the curriculum.

      Results

       Cognitive Assessment

      The pre-curriculum mean test score was 5.4 (SD, 2.61), which increased to 7 (SD, 0.89) for the post-test (P = .04). This indicates overall knowledge improvement and retention of knowledge gained during the course of a curriculum spanning several months.

       Simulations (Resections)

      Volume of resection was recorded for the initial, usually largest, specimen, as well as the total volume resected for each model. Overall, across all 5 models, initial volumes were significantly different between faculty and residents (Figure 2; mean difference, 2.5 cm3; SD, 0.36 cm3; P < .0001). Similarly for the total volumes there was a significant difference between faculty and residents (Figure 3; mean difference, 8.6 cm3; SD, 1.3 cm3; P < .0001). Residents resected increasingly larger volumes of tissue over the course of the study. Initial resection volume (compared with faculty) increased by 0.82 cm3 per resection (Figure 4; 95% confidence interval [CI], 0.37-1.3 cm3; P < .0001). Total resection volume (compared with faculty) increased by 3.6 cm3 per resection (Figure 5; 95% CI, 2.4-4.9 cm3; P < .0001). This difference is despite the fact that the faculty resection volume was the greatest for model 1 and the least for model 5 just on the basis of the differing patient morphology and random order of clinical presentation.
      Figure thumbnail gr2
      Figure 2Plot showing volumes of “initial” resection specimen for each 3-D printed model. A circle surrounds the faculty data point.
      Figure thumbnail gr3
      Figure 3Plot showing volumes of “total” resection specimen for each 3-D printed model. A circle surrounds the faculty data point.
      Figure thumbnail gr4
      Figure 4Plot showing change in volume of “initial” resection specimen for residents (compared with faculty) over the course of the study. Each colored line represents a resident during the study. The red zero line corresponds to the faculty resection volume.
      Figure thumbnail gr5
      Figure 5Plot showing change in volume of “total” resection specimen for residents (compared with faculty) over the course of the study. Each colored line represents a resident during the study. The red zero line corresponds to the faculty resection volume.
      Agreement between the simulated and patient resection volumes was similar to that reported in our previous experience.
      • Hermsen J.L.
      • Burke T.M.
      • Seslar S.P.
      • Owens D.S.
      • Ripley B.A.
      • Mokadam N.A.
      • et al.
      Scan, plan, print, practice, perform: development and use of a patient-specific 3-dimensional printed model in adult cardiac surgery.
      No patient left the operating room with an LVOT velocity >2 m/s.

       Simulations (Video Assessments)

      Agreement between faculty assessments was poor (Figure 6; P < .0001) with rater 3 having significantly higher ratings compared with rater 1 (−0.93; 95% CI, −1.3 to −0.58) and rater 2 (−0.9; 95% CI, −1.3 to −0.55). Excluding the faculty member who rated himself, who was also the sole faculty member performing the model resections, faculty assessment scores did not identify faculty status (P = .72).
      Figure thumbnail gr6
      Figure 6Three-dimensional box plot showing poor correlation between video scores assigned by 3 faculty reviewers.

       Survey

      Survey responses were categorized into values between 1 (poor) and 5 (excellent) for all domains. The median response was 4 (interquartile range, 4-5) and response mean was 4.31 (SD, 0.66; standard error, 0.67) indicating an overall favorable assessment. Table 1 shows a summary of survey responses for domains relating to physical attributes of the prints and the realism of the simulation experience. Table E1, Table E2, Table E3, Table E4 show survey results for other domains.
      Table 1Survey responses for domains relating to physical attributes of the prints and the realism of the simulation experience
      DomainDon't knowNot at all realisticMildly realistic, but not enough for this useModerately realistic, enough for this useHighly realistic, enough for this use
      Physical attributes
       Chest depth100
       Landmark visualization
      RCA os6040
      Annulus8020
      Membranous septum2080
      Right trigone100
      Left trigone100
       Septal visualization6040
      Realism of materials
       Tissue quality4060
      Realism of experience
       Realism of surgical approach206020
       Realism of surgical angles8020
       Realism of ability to manipulate tissue2080
       Realism of response to scalpel4060
       Realism of septal anatomy6040
       Realism of overall experience206020
      Data are presented as percentages. RCA os, Right coronary artery ostium.

      Discussion

      Three-dimensional printing in surgical specialties continues to evolve and has been applied in multiple specialties. Its use in cardiac surgery has been largely focused on congenital disease because the anatomy is often unusual, and there is a limited time to visualize, decipher, and make decisions while the patient's heart is arrested.
      • Oliveiri L.J.
      • Su L.
      • Hynes C.F.
      • Krieger A.
      • Alfares F.A.
      • Ramakrishnan K.
      • et al.
      “Just-in-time” simulation training using 3-D printed cardiac models after congenital cardiac surgery.
      • Yoo S.J.
      • Spray T.
      • Austin E.H.
      • Yun T.J.
      • van Arsdell G.
      Hands-on surgical training of congenital heart surgery using 3-dimensional print models.
      The advancement in print media and anatomic detail over the past 5 years has been substantial.
      • Yoo S.J.
      • Spray T.
      • Austin E.H.
      • Yun T.J.
      • van Arsdell G.
      Hands-on surgical training of congenital heart surgery using 3-dimensional print models.
      • Shiraishi I.
      • Yamagishi M.
      • Hamaoka K.
      • Fukuzawa M.
      • Yagihara T.
      Simulative operation on congenital heart disease using rubber-like urethane stereolithographic biomodels based on 3-D datasets of multislice computed tomography.
      Other surgical specialties have similarly begun to use 3-D printing for simulation, or surgical rehearsal, of uncommon or unusual procedures because of the known relationships between volume and outcome.
      • Mamidanna R.
      • Ni Z.
      • Anderson O.
      • Spiegelhalter S.D.
      • Botle A.
      • Aylin P.
      • et al.
      Surgeon volume and cancer esophagectomy, gastrectomy, and pancreatectomy: a population based study in England.
      Neurosurgeons have developed high-fidelity 3-D printed models to simulate aneurysm clipping, and have shown agreement in clip selection between simulations and actual procedures.
      • Wang L.
      • Ye X.
      • Hao Q.
      • Chen Y.
      • Chen X.
      • Wang H.
      • et al.
      Comparison of two three-dimensional printed models of complex intracranial aneurysms for surgical simulation.
      Orthopedists have shown simulation of tibial plateau fracture repair using a patient-specific 3-D print to decrease surgery times.
      • Lou Y.
      • Cai L.
      • Wang C.
      • Tang Q.
      • Pan T.
      • Guo X.
      • et al.
      Comparison of traditional surgery and surgery assisted by three dimensional printing technology in the treatment of tibial plateau fractures.
      The hypothesis that residents are undertaught in this procedure was generally confirmed with the results of the first simulation of the series, which served as a baseline assessment. Uniformly, the residents resected a very small total volume of septal tissue (2.5-4.0 cm3). This would have constituted a sham procedure, because the actual volume resected at surgery was greater by at least 4-fold (15 cm3; Figure 7, left image). The resident resections for patient 1 were generally inadequate even for all other patients considering faculty resection volumes of 17, 9.5, 10.5, and 5.5 cm3 for the 4 other models, respectively. However, although the resection volumes in aggregate (for all study patients) were significantly different between attending and resident, the residents clearly made progress with accumulation of experience performing myectomy on the models as shown by significant increases in the volumes of resection.
      Figure thumbnail gr7
      Figure 7Images of 3-D printed models after myectomy by faculty and residents. The left panel shows faculty and resident models for patient 1 and the right panel shows faculty and resident models for patient 4.
      Importantly, the subjective qualities (shape, depth, apical extent) of the resident resections also improved during the study and were much more on par with the attending resection at the end of the study (Figure 7, right image). This is significant because the objective of the study was only to perform basic evaluation of the model as a platform for education. Therefore, residents were not formally instructed or coached during their model resections. Either purely visual (comparison of attending and resident models for patient 1 before working on patient 2) or relatively generic feedback, such as “create an edge of specimen to retract posteriorly then flatten the knife angle to proceed more apically,” was given during the study. We speculate that the act of doing the model resections provided the residents a more educated lens through which to view and mentally engage during each subsequent model, creating a type of positive feedback loop, which translated to improving resections as the study progressed. This simulation platform seems poised, with more focused and real-time coaching and feedback, to be an excellent tool for instruction in extended septal myectomy.
      As discussed in a previous related publication, many porcine hearts used in cardiac surgery simulation curricula possess asymmetric basilar septal hypertrophy that might allow for meaningful myectomy.
      • Hermsen J.L.
      • Burke T.M.
      • Seslar S.P.
      • Owens D.S.
      • Ripley B.A.
      • Mokadam N.A.
      • et al.
      Scan, plan, print, practice, perform: development and use of a patient-specific 3-dimensional printed model in adult cardiac surgery.
      However, the internal dimensions, working space, and angles afforded by these hearts would likely not translate well to human patients. Additionally, the patient-specific nature afforded by the 3-D printed models would be lost. In a disease with such variable relevant morphology, the patient-specific nature of a simulation or surgical rehearsal is important. Even with only 5 patients, the variability of phenotype in hypertrophic cardiomyopathy was on display. In each case, a very similar section of heart and aorta was printed, and these print volumes ranged from 137 cm3 to 289 cm3.
      The results of the survey (Table 1 and Table E1, Table E2, Table E3, Table E4) were encouraging. Although the materials used to produce the prints can definitely be improved and become more tissue-like, most of the residents thought it adequate for this purpose. The “face validity” of the simulator was also judged favorably on the basis of the realism scores related to recreation of the surgical approach and working angles. Most importantly, most of the residents thought the simulations were worth the time and effort expended (approximately 30 minutes for each model, usually in the midst of a busy clinical day) and would enable them to perform myectomy with senior partner assistance when entering practice. Additionally, each resident benefited from the anatomy of 5 different patients, by performing simulations on each of the 5 models, although they each operated on only 1 patient.

       Limitations

      This is a relatively small study in number of patients and residents. Cost is currently a barrier for more widespread use of this technology. For this study, 30 models were printed at a total cost of approximately $8000 USD (approximately $270 per model). The print material technology is still suboptimal as was noted by our residents and has been noted by others using 3-D cardiac prints for surgical simulation.
      • Yoo S.J.
      • Spray T.
      • Austin E.H.
      • Yun T.J.
      • van Arsdell G.
      Hands-on surgical training of congenital heart surgery using 3-dimensional print models.

       Future Directions

      This experience represents a small “library” of cases because an adequate patient resection volume, inferred from excellent surgical results, is known. Such a series of cases could easily be reprinted as a “case series” for another learner to work through. If paired with surgical videos (an answer key of sorts), a learner could perform a resection, measure the volumes, and then review the video for comparison with the actual patient outcome. This curriculum could quite easily be integrated into established cardiac surgical simulation programs as a learning module, similar to existing modules focused on aortic valve replacement or coronary artery bypass grafting (CABG).
      • Feins R.H.
      • Burkhart H.M.
      • Conte J.V.
      • Coore D.N.
      • Fann J.I.
      • Hicks J.L.
      • et al.
      Simulation-based training in cardiac surgery.
      Additionally this model could provide feedback regarding excessive resection and creation of a ventricular septal defect (VSD). The model is not well equipped in the current configuration to simulate VSD repair because the right atrium and tricuspid valve are not included in the print.

      Conclusions

      The ability for residents to engage in deliberate practice on 3-D prints representing “real anatomy,” and to compare their resection to one done by an attending surgeon on an identical print might provide a new paradigm to enable improved training in myectomy surgery. Because of the difficulty in teaching this procedure to residents with the “see one, do one, teach one” model, this might help to fill an important training gap in most resident's surgical repertoire. Although 3-D printing technology is not yet sufficient to provide meaningful simulation for more commonly performed CABG or valve procedures, it is well suited, and might find a niche as a simulation platform for septal myectomy.

       Conflict of Interest Statement

      Dr Hermsen's spouse is a paid consultant for Medtronic. Dr Burke is the owner of a for profit corporation that manufactures patient-specific heart models and received reimbursement for materials used in producing heart models for this study. Dr Mokadam is a consultant for Medtronic and Abbott, and an investigator for Medtronic, Abbott, and Syncardia. All other authors have nothing to disclose with regard to commercial support.

      Appendix E1. Pre- and post-test cognitive assessment

      Figure thumbnail fx3

      Appendix E2

      Figure thumbnail fx4
      Table E1Survey results for self-rating using the simulator
      Almost impossible to performVery difficult to performDifficult to performSomewhat easy to performVery easy to perform
      Please rate your ability to perform the below tasks on this simulator
       Initial resection20%60%20%
       Subsequent resection60%20%20%
       Palpation of septal thickness60%40%
      Table E2Survey results for readiness for educational purpose
      Requires complete redesignRequires major changes before useRequires small changes before useCan be used as is
      Readiness assessment for educational purpose80%20%
      Table E3Survey results for value of simulator as a teaching tool and relevance to practice
      Don't knowNo valueLittle valueModerate valueGreat deal of value
      Please rate the value of the simulator as a teaching tool60%40%
      Relevance to practice, assuming employed as adult cardiac surgeon20%60%20%
      Table E4Survey results for value added
      No valueLittle value, not worth the effortModest value, worth the effortHigh value, worth the effort; would perform myectomy with senior partnerHigh value, worth the effort; would perform myectomy independently
      Value added
       How would you rate this experience with regard to your comfort in performing myectomy in the future100%

      Supplementary Data

      References

        • Maron B.J.
        • Rastegar H.
        • Udelson J.E.
        • Dearani J.A.
        • Maron M.S.
        Contemporary surgical management of hypertrophic cardiomyopathy, the need for more myectomy surgeons and disease-specific centers, and the Tufts initiative.
        Am J Cardiol. 2013; 112: 1512-1515
        • Hermsen J.L.
        • Burke T.M.
        • Seslar S.P.
        • Owens D.S.
        • Ripley B.A.
        • Mokadam N.A.
        • et al.
        Scan, plan, print, practice, perform: development and use of a patient-specific 3-dimensional printed model in adult cardiac surgery.
        J Thorac Cardiovasc Surg. 2017; 153: 132-140
        • Mokadam N.A.
        • Fann J.I.
        • Hicks G.L.
        • Nesbitt J.C.
        • Burkhart H.M.
        • Conte J.V.
        • et al.
        Experience with the cardiac surgery simulation curriculum: results of the resident and faculty survey.
        Ann Thorac Surg. 2017; 103: 322-328
        • Cook M.R.
        • Graff-Baker A.N.
        • Moren A.M.
        • Brown S.
        • Fair K.A.
        • Kiraly L.N.
        • et al.
        A disease-specific hybrid rotation increases opportunities for deliberate practice.
        J Surg Educ. 2016; 73: 1-6
        • Smedira N.G.
        • Lytle B.W.
        • Lever H.M.
        • Rajeswaran J.
        • Krishnaswamy G.
        • Kaple R.K.
        • et al.
        Current effectiveness and risks of isolated septal myectomy for hypertrophic obstructive cardiomyopathy.
        Ann Thorac Surg. 2008; 85: 127-134
        • Maron B.J.
        • Ommen S.R.
        • Semsarian C.
        • Spirito P.
        • Olivotto I.
        • Maron M.S.
        Hypertrophic cardiomyopathy: present and future, with translation into contemporary cardiovascular medicine.
        J Am Coll Cardiol. 2014; 64: 83-99
        • Schaff H.V.
        • Said S.
        Transaortic extended septal myectomy for hypertrophic cardiomyopathy.
        Op Tech Thor Cardiovasc Surg. 2012; 17: 238-250
        • YouTube
        Video of my heart surgery (myectomy).
        (Available at:) (Accessed July 3, 2018)
        • Barsness K.A.
        • Rooney D.M.
        • Davis L.M.
        Collaboration in simulation: the development and initial validation of a novel thoracoscopic neonatal simulator.
        J Pediatr Surg. 2013; 48: 1232-1238
        • Oliveiri L.J.
        • Su L.
        • Hynes C.F.
        • Krieger A.
        • Alfares F.A.
        • Ramakrishnan K.
        • et al.
        “Just-in-time” simulation training using 3-D printed cardiac models after congenital cardiac surgery.
        World J Pediatr Congenit Heart Surg. 2016; 7: 164-168
        • Yoo S.J.
        • Spray T.
        • Austin E.H.
        • Yun T.J.
        • van Arsdell G.
        Hands-on surgical training of congenital heart surgery using 3-dimensional print models.
        J Thorac Cardiovasc Surg. 2017; 153: 1530-1540
        • Shiraishi I.
        • Yamagishi M.
        • Hamaoka K.
        • Fukuzawa M.
        • Yagihara T.
        Simulative operation on congenital heart disease using rubber-like urethane stereolithographic biomodels based on 3-D datasets of multislice computed tomography.
        Eur J Cardiothorac Surg. 2010; 37: 302-306
        • Mamidanna R.
        • Ni Z.
        • Anderson O.
        • Spiegelhalter S.D.
        • Botle A.
        • Aylin P.
        • et al.
        Surgeon volume and cancer esophagectomy, gastrectomy, and pancreatectomy: a population based study in England.
        Ann Surg. 2016; 263: 727-732
        • Wang L.
        • Ye X.
        • Hao Q.
        • Chen Y.
        • Chen X.
        • Wang H.
        • et al.
        Comparison of two three-dimensional printed models of complex intracranial aneurysms for surgical simulation.
        World Neurosurg. 2017; 103: 671-679
        • Lou Y.
        • Cai L.
        • Wang C.
        • Tang Q.
        • Pan T.
        • Guo X.
        • et al.
        Comparison of traditional surgery and surgery assisted by three dimensional printing technology in the treatment of tibial plateau fractures.
        Int Orthop. 2017; 41: 1875-1880
        • Feins R.H.
        • Burkhart H.M.
        • Conte J.V.
        • Coore D.N.
        • Fann J.I.
        • Hicks J.L.
        • et al.
        Simulation-based training in cardiac surgery.
        Ann Thorac Surg. 2017; 103: 312-321
      1. Feins R.H. Burkhart H.M. Coore D.N. Conte J.V. Fann J.I. Hicks G.L. Cardiac Surgery Simulation Curriculum. Lulu, Chapel Hill, NC2015