- Email: [email protected]
Novel Application of Rapid Prototyping for Simulation of Bronchoscopic Anatomy
EMERGING TECHNOLOGY REVIEW Gerard R. Menecke, Jr, MD Marco Ranucci, MD Section Editors
Novel Application of Rapid Prototyping for Simulation of Bronchoscopic Anatomy Sergio Bustamante, MD,* Somnath Bose, MBBS, MD,† Paul Bishop, MSEE, RVT,‡ Ryan Klatte, BSBME,§ and Frederick Norris, MD* Objective: The authors used rapid prototyping (RP) technology to create anatomically congruent models of tracheo-bronchial tree for teaching relevant bronchoscopic anatomy. Design: Pilot study. Setting: A single level tertiary academic medical center. Interventions: Two 3 dimensional (3D) models of tracheobronchial tree (one showing normal anatomy and another with an early take off of right apical bronchus) were recreated from Computed Tomographic images using RP technology. These images were then attached to mannequins and examined with a flexible fiberoptic bronchoscope
(FFB). These images were then compared with the actual FFB images obtained during lung isolation. Measurements and Main Results: The images obtained through the 3D models were found to be congruent to actual patient anatomy. Conclusions: RP can be successfully used to create anatomically accurate models from imaging studies. There is potential for RP to become a valuable educational tool in the future. & 2014 Elsevier Inc. All rights reserved.
R
accurate models of the tracheobronchial tree for teaching tracheobronchial anatomy.
APID PROTOTYPING (RP) is a generic term that comprises several technologies, including laser sintering, stereolithography, fused deposition modeling, and threedimensional (3D) printing.1 Since its introduction more than 20 years ago,2 it has found increasing application for simulation of complex structures in different medical subspecialties.3–6 It already has been used in simulation of cranio-maxillofacial surgery,3 surgeries for abdominal aortic aneurysm,4 transcatheter cardiac valve placement,5 and in planning of neurosurgical procedures.6 RP has evolved into a useful teaching tool, because it can be uniquely customized according to a patient’s anatomic and pathologic characteristics. A search of relevant peerreviewed literature in the English language did not yield any article focusing on application of RP in the field of anesthesia. This brief review serves to illustrate the potential application of this relatively new technology in fabricating anatomically
KEY WORDS: rapid prototyping, three dimensional printing, bronchoscopic anatomy, simulation in anesthesia, lung isolation, anesthesia education
BACKGROUND
It has been recognized that anesthesiologists with limited thoracic experience face difficulties in obtaining lung isolation. This is attributed to insufficient knowledge of bronchial anatomy and lack of proficiency with flexible bronchoscopy (FB).7 Current educational tools used to teach bronchial anatomy and FB include online simulation, instructional DVDs, mannequin simulators, and virtual reality simulators.8 The use of RP is a novel concept in anesthesia simulation. The authors used this technology to create 2 anatomically accurate 3D models of the tracheobronchial tree for teaching the relevant anatomy. METHODS
From the *Department of Cardiothoracic Anesthesiology, Cleveland Clinic Foundation, Cleveland, OH; †Anesthesiology Institute, Cleveland Clinic Foundation, Cleveland, OH; ‡Department of Vascular Surgery, Peripheral Core Lab, Cleveland Clinic Foundation, Cleveland, OH; and §Lerner Research Institute, Cleveland Clinic Foundation, Cleveland, OH. * Presented in part at the 2013 Annual Meeting of the International Anesthesia Research Society, San Francisco, May 4 to May 7, 2013. Address reprint requests to Somnath Bose, MD, Anesthesiology Institute, Cleveland Clinic Foundation, 9500 Euclid Avenue, E31, Cleveland, OH 44195. E-mail: [email protected] © 2014 Elsevier Inc. All rights reserved. 1053-0770/2601-0001$36.00/0 http://dx.doi.org/10.1053/j.jvca.2013.08.015 1122
The authors used computed tomographic images of the thorax that were in the DICOM (Digital Imaging and Communication in Medicine) format to create the models; however, magnetic resonance images also could be used. The lung window series was 3D-reconstructed using Tera-Recon Aquarius Intuition (TeraRecon, Foster City, CA) software. A semiautomated process then selectively identified the area of interest —in this case, the tracheobronchial tree. These images then were converted into a Standard Tesselation Language (STL) file format and additional digital postprocessing was performed (Fig 1). The STL format describes the surface geometry and spatial orientation of a 3D object, which can be interpreted by a compatible 3D printer. Finally, this file was sent to the 3D printer (Fig 2 shows the CT scan and the 3D rendering of one of the models before printing). The 3D printer works similarly to an inkjet printer, depositing ultraviolet (UV) light-hardened liquid resin on a platform in a pattern that corresponds to the digital file provided. The printing is done in
Journal of Cardiothoracic and Vascular Anesthesia, Vol 28, No 4 (August), 2014: pp 1122–1125
1123
SIMULATION OF BRONCHOSCOPIC ANATOMY
Fig 1. Flowchart showing technique of rapid prototyping as applied to tracheobronchial models. Abbreviations: 3D, three-dimensional; CT, computed tomography; DICOM, Digital Imaging and Communication in Medicine; MRI, magnetic resonance imaging. (Color version of figure is available online.)
multiple layers, with each layer adding height to the model. Temporary support material fills in the overhanging cavities or geometry. The authors used an Object Connex 350 (Stratasys, Inc., Eden Prairie, MN) printer and a photosensitive flexible liquid resin, which provided a more realistic feel to the prototypes. The process of printing took approximately 7 to 10 hours. The model then was dyed and painted appropriately to give it an anatomically correct representation. The authors used two CT scans of the thorax to create two 3D models as a proof of concept—the concept being that 3D prototypes potentially could be useful in teaching both normal and abnormal anatomy of the tracheobronchial tree, which eventually would lead to improved knowledge of tracheobronchial anatomy and lung isolation skills among anesthesiology residents. However, the authors believe that the application
of this technology can be expanded to educate practitioners at any level, including bronchoscopists and pulmonologists. The first CT scan was that of a normal tracheobronchial tree, and the second CT scan was of a bronchial tree with an early take-off of the right apical bronchus that had posed difficulties during lung isolation. The models constructed by RP then were connected to a mannequin and were examined with a flexible fiberoptic bronchoscope (FFB). The bronchoscopic images then were compared with the actual FB views obtained during lung isolation (Fig 3 shows the model and FFB views of the normal tracheobronchial tree; Fig 4 shows the 3D-generated perspective image of the early take-off of the right apical bronchus and the comparison of FB images of the model and the actual anatomy). Comparison showed that both images were similar. The bronchoscopic image of the 3D model was found to be congruent with the patient’s anatomy. The quality of RP is dependent on the imaging available; the authors used a 2.5-mm and a 3-mm slice thickness CT scan for the prototypes. The prototypes would have been more realistic if a higher resolution CT scan had been used to create the models. DISCUSSION
The use of RP is a relatively new concept in medical education. As is evident from this pilot study, this technology can be used in simulation in anesthesia education. RP takes advantage of air and soft tissue contrast on a CT scan or by magnetic resonance imaging, which clearly show the bronchial anatomy. After adequate post-processing, the obtained information then can be exported to a 3D printer that allows for highfidelity reconstruction. Because unfamiliarity with endoscopic bronchial anatomy is considered to be an important factor in failure to achieve lung isolation, the authors believe that the accurate representation of anatomical structures using this
Fig 2. (Top left) Coronal CT scan showing early takeoff of right apical bronchus (red arrow). (Color version of figure is available online.)(Top right) Computer-generated 3D image of the tracheobronchial tree, showing the early take-off of the bronchus. (Bottom right) 3D model of the tracheobronchial tree. (Color version of figure is available online.)
1124
BUSTAMANTE ET AL
Fig 3. (Top left) 3D rendering of a normal tracheobronchial tree. (Top right) Fiberoptic view of the 3D model through the bronchus intermedius. (Bottom right) Fiberoptic view of the actual anatomy through the bronchus intermedius. Note the similarity between the top right and the bottom right views. (Color version of figure is available online.)
Fig 4. (Top left) CT-generated 3D perspective view. (Top right) Fiberoptic view of original anatomy. (Bottom right) Fiberoptic view through 3D model. Red arrow indicated early take-off of the right apical bronchus. Blue arrow indicates bronchial cuff of 35Fr left double-lumen tube. (Color version of figure is available online.)
1125
SIMULATION OF BRONCHOSCOPIC ANATOMY
technology potentially could be helpful in bridging that gap.7 This technology can be used to recreate relatively uncommon anatomic variants of the tracheobronchial tree that otherwise would be difficult to replicate in a training environment. The application of this technology could be expanded to create realistic, anatomically congruent models of difficult airway scenarios of almost any perceivable complexity, which then could be used as an invaluable tool in the teaching and training of management of such scenarios. Thus, this technology potentially might find application in the actual preoperative evaluation and preparation for patients with challenging airway anatomy. Examples of such scenarios include a thoracic aortic aneurysm, a tumor impinging on the trachea, or a distorted upper airway.
CONCLUSION
Because this technology is relatively nascent, the biggest limiting factor is the cost involved in the construction; another limiting factor is the availability of the technology. The estimated cost of materials is approximately $250 per model. This cost is comparable, if not less than, many mannequin simulators and much less expensive than virtual reality simulators. The reproduction of the structures is dependent on the resolution of the images available—the higher the resolution, the better the reproduction. In conclusion, continued exploration of rapid prototyping is needed, and it seems that there is potential for it to become a valuable tool for education in the future.
REFERENCES 1. A factory on your desk. Manufacturing: Producing solid objects, even quite complex ones, with 3D printers is gradually becoming easier and cheaper. The Economist Technology Quarterly 3, September 2009. Accessed March 15, 2013. 2. Jones N: Science in three dimensions: The print revolution. Nature 487:22-23, 2012 3. Waran V, Menon R, Pancharatnam D, et al: The creation and verification of cranial models using three-dimensional rapid prototyping technology in field of transnasal sphenoid endoscopy. Am J Rhinol Allergy 26:132-136, 2012 4. Wilasrusmee C, Suvikrom J, Suthakorn J, et al: Three-dimensional aortic aneurysm model and endovascular repair: An educational tool for surgical trainees. Int J Angiol 17:129-133, 2008 5. Schmauss D, Schmitz C, Bigdeli AK, et al: Three-dimensional printing of models for preoperative planning and simulation
of transcatheter valve placement. Ann Thorac Surg 93:31-33, 2012 6. Spottiswoode BS, van den Heever DJ, Chang Y, et al: Preoperative three-dimensional model creation of magnetic resonance brain images as a tool to assist neurosurgical planning. Stereotact Funct Neurosurg 91:162-169, 2013 7. Campos JH, Hallam EA, Van Natta T, et al: Devices for lung isolation used by anesthesiologists with limited thoracic experience: Comparison of double-lumen endotracheal tube, Univent torque control blocker, and Arndt wire-guided endobronchial blocker. Anesthesiology 104:261-266, 2006 8. Campos JH, Hallam EA, Ueda K: Training in placement of the left-sided double-lumen tube among non-thoracic anaesthesiologists: Intubation model simulator versus computer-based digital video disc, a randomised controlled trial. Eur J Anaesthesiol 28:169-174, 2011
Novel Application of Rapid Prototyping for Simulation of Bronchoscopic Anatomy Sergio Bustamante, MD,* Somnath Bose, MBBS, MD,† Paul Bishop, MSEE, RVT,‡ Ryan Klatte, BSBME,§ and Frederick Norris, MD* Objective: The authors used rapid prototyping (RP) technology to create anatomically congruent models of tracheo-bronchial tree for teaching relevant bronchoscopic anatomy. Design: Pilot study. Setting: A single level tertiary academic medical center. Interventions: Two 3 dimensional (3D) models of tracheobronchial tree (one showing normal anatomy and another with an early take off of right apical bronchus) were recreated from Computed Tomographic images using RP technology. These images were then attached to mannequins and examined with a flexible fiberoptic bronchoscope
(FFB). These images were then compared with the actual FFB images obtained during lung isolation. Measurements and Main Results: The images obtained through the 3D models were found to be congruent to actual patient anatomy. Conclusions: RP can be successfully used to create anatomically accurate models from imaging studies. There is potential for RP to become a valuable educational tool in the future. & 2014 Elsevier Inc. All rights reserved.
R
accurate models of the tracheobronchial tree for teaching tracheobronchial anatomy.
APID PROTOTYPING (RP) is a generic term that comprises several technologies, including laser sintering, stereolithography, fused deposition modeling, and threedimensional (3D) printing.1 Since its introduction more than 20 years ago,2 it has found increasing application for simulation of complex structures in different medical subspecialties.3–6 It already has been used in simulation of cranio-maxillofacial surgery,3 surgeries for abdominal aortic aneurysm,4 transcatheter cardiac valve placement,5 and in planning of neurosurgical procedures.6 RP has evolved into a useful teaching tool, because it can be uniquely customized according to a patient’s anatomic and pathologic characteristics. A search of relevant peerreviewed literature in the English language did not yield any article focusing on application of RP in the field of anesthesia. This brief review serves to illustrate the potential application of this relatively new technology in fabricating anatomically
KEY WORDS: rapid prototyping, three dimensional printing, bronchoscopic anatomy, simulation in anesthesia, lung isolation, anesthesia education
BACKGROUND
It has been recognized that anesthesiologists with limited thoracic experience face difficulties in obtaining lung isolation. This is attributed to insufficient knowledge of bronchial anatomy and lack of proficiency with flexible bronchoscopy (FB).7 Current educational tools used to teach bronchial anatomy and FB include online simulation, instructional DVDs, mannequin simulators, and virtual reality simulators.8 The use of RP is a novel concept in anesthesia simulation. The authors used this technology to create 2 anatomically accurate 3D models of the tracheobronchial tree for teaching the relevant anatomy. METHODS
From the *Department of Cardiothoracic Anesthesiology, Cleveland Clinic Foundation, Cleveland, OH; †Anesthesiology Institute, Cleveland Clinic Foundation, Cleveland, OH; ‡Department of Vascular Surgery, Peripheral Core Lab, Cleveland Clinic Foundation, Cleveland, OH; and §Lerner Research Institute, Cleveland Clinic Foundation, Cleveland, OH. * Presented in part at the 2013 Annual Meeting of the International Anesthesia Research Society, San Francisco, May 4 to May 7, 2013. Address reprint requests to Somnath Bose, MD, Anesthesiology Institute, Cleveland Clinic Foundation, 9500 Euclid Avenue, E31, Cleveland, OH 44195. E-mail: [email protected] © 2014 Elsevier Inc. All rights reserved. 1053-0770/2601-0001$36.00/0 http://dx.doi.org/10.1053/j.jvca.2013.08.015 1122
The authors used computed tomographic images of the thorax that were in the DICOM (Digital Imaging and Communication in Medicine) format to create the models; however, magnetic resonance images also could be used. The lung window series was 3D-reconstructed using Tera-Recon Aquarius Intuition (TeraRecon, Foster City, CA) software. A semiautomated process then selectively identified the area of interest —in this case, the tracheobronchial tree. These images then were converted into a Standard Tesselation Language (STL) file format and additional digital postprocessing was performed (Fig 1). The STL format describes the surface geometry and spatial orientation of a 3D object, which can be interpreted by a compatible 3D printer. Finally, this file was sent to the 3D printer (Fig 2 shows the CT scan and the 3D rendering of one of the models before printing). The 3D printer works similarly to an inkjet printer, depositing ultraviolet (UV) light-hardened liquid resin on a platform in a pattern that corresponds to the digital file provided. The printing is done in
Journal of Cardiothoracic and Vascular Anesthesia, Vol 28, No 4 (August), 2014: pp 1122–1125
1123
SIMULATION OF BRONCHOSCOPIC ANATOMY
Fig 1. Flowchart showing technique of rapid prototyping as applied to tracheobronchial models. Abbreviations: 3D, three-dimensional; CT, computed tomography; DICOM, Digital Imaging and Communication in Medicine; MRI, magnetic resonance imaging. (Color version of figure is available online.)
multiple layers, with each layer adding height to the model. Temporary support material fills in the overhanging cavities or geometry. The authors used an Object Connex 350 (Stratasys, Inc., Eden Prairie, MN) printer and a photosensitive flexible liquid resin, which provided a more realistic feel to the prototypes. The process of printing took approximately 7 to 10 hours. The model then was dyed and painted appropriately to give it an anatomically correct representation. The authors used two CT scans of the thorax to create two 3D models as a proof of concept—the concept being that 3D prototypes potentially could be useful in teaching both normal and abnormal anatomy of the tracheobronchial tree, which eventually would lead to improved knowledge of tracheobronchial anatomy and lung isolation skills among anesthesiology residents. However, the authors believe that the application
of this technology can be expanded to educate practitioners at any level, including bronchoscopists and pulmonologists. The first CT scan was that of a normal tracheobronchial tree, and the second CT scan was of a bronchial tree with an early take-off of the right apical bronchus that had posed difficulties during lung isolation. The models constructed by RP then were connected to a mannequin and were examined with a flexible fiberoptic bronchoscope (FFB). The bronchoscopic images then were compared with the actual FB views obtained during lung isolation (Fig 3 shows the model and FFB views of the normal tracheobronchial tree; Fig 4 shows the 3D-generated perspective image of the early take-off of the right apical bronchus and the comparison of FB images of the model and the actual anatomy). Comparison showed that both images were similar. The bronchoscopic image of the 3D model was found to be congruent with the patient’s anatomy. The quality of RP is dependent on the imaging available; the authors used a 2.5-mm and a 3-mm slice thickness CT scan for the prototypes. The prototypes would have been more realistic if a higher resolution CT scan had been used to create the models. DISCUSSION
The use of RP is a relatively new concept in medical education. As is evident from this pilot study, this technology can be used in simulation in anesthesia education. RP takes advantage of air and soft tissue contrast on a CT scan or by magnetic resonance imaging, which clearly show the bronchial anatomy. After adequate post-processing, the obtained information then can be exported to a 3D printer that allows for highfidelity reconstruction. Because unfamiliarity with endoscopic bronchial anatomy is considered to be an important factor in failure to achieve lung isolation, the authors believe that the accurate representation of anatomical structures using this
Fig 2. (Top left) Coronal CT scan showing early takeoff of right apical bronchus (red arrow). (Color version of figure is available online.)(Top right) Computer-generated 3D image of the tracheobronchial tree, showing the early take-off of the bronchus. (Bottom right) 3D model of the tracheobronchial tree. (Color version of figure is available online.)
1124
BUSTAMANTE ET AL
Fig 3. (Top left) 3D rendering of a normal tracheobronchial tree. (Top right) Fiberoptic view of the 3D model through the bronchus intermedius. (Bottom right) Fiberoptic view of the actual anatomy through the bronchus intermedius. Note the similarity between the top right and the bottom right views. (Color version of figure is available online.)
Fig 4. (Top left) CT-generated 3D perspective view. (Top right) Fiberoptic view of original anatomy. (Bottom right) Fiberoptic view through 3D model. Red arrow indicated early take-off of the right apical bronchus. Blue arrow indicates bronchial cuff of 35Fr left double-lumen tube. (Color version of figure is available online.)
1125
SIMULATION OF BRONCHOSCOPIC ANATOMY
technology potentially could be helpful in bridging that gap.7 This technology can be used to recreate relatively uncommon anatomic variants of the tracheobronchial tree that otherwise would be difficult to replicate in a training environment. The application of this technology could be expanded to create realistic, anatomically congruent models of difficult airway scenarios of almost any perceivable complexity, which then could be used as an invaluable tool in the teaching and training of management of such scenarios. Thus, this technology potentially might find application in the actual preoperative evaluation and preparation for patients with challenging airway anatomy. Examples of such scenarios include a thoracic aortic aneurysm, a tumor impinging on the trachea, or a distorted upper airway.
CONCLUSION
Because this technology is relatively nascent, the biggest limiting factor is the cost involved in the construction; another limiting factor is the availability of the technology. The estimated cost of materials is approximately $250 per model. This cost is comparable, if not less than, many mannequin simulators and much less expensive than virtual reality simulators. The reproduction of the structures is dependent on the resolution of the images available—the higher the resolution, the better the reproduction. In conclusion, continued exploration of rapid prototyping is needed, and it seems that there is potential for it to become a valuable tool for education in the future.
REFERENCES 1. A factory on your desk. Manufacturing: Producing solid objects, even quite complex ones, with 3D printers is gradually becoming easier and cheaper. The Economist Technology Quarterly 3, September 2009. Accessed March 15, 2013. 2. Jones N: Science in three dimensions: The print revolution. Nature 487:22-23, 2012 3. Waran V, Menon R, Pancharatnam D, et al: The creation and verification of cranial models using three-dimensional rapid prototyping technology in field of transnasal sphenoid endoscopy. Am J Rhinol Allergy 26:132-136, 2012 4. Wilasrusmee C, Suvikrom J, Suthakorn J, et al: Three-dimensional aortic aneurysm model and endovascular repair: An educational tool for surgical trainees. Int J Angiol 17:129-133, 2008 5. Schmauss D, Schmitz C, Bigdeli AK, et al: Three-dimensional printing of models for preoperative planning and simulation
of transcatheter valve placement. Ann Thorac Surg 93:31-33, 2012 6. Spottiswoode BS, van den Heever DJ, Chang Y, et al: Preoperative three-dimensional model creation of magnetic resonance brain images as a tool to assist neurosurgical planning. Stereotact Funct Neurosurg 91:162-169, 2013 7. Campos JH, Hallam EA, Van Natta T, et al: Devices for lung isolation used by anesthesiologists with limited thoracic experience: Comparison of double-lumen endotracheal tube, Univent torque control blocker, and Arndt wire-guided endobronchial blocker. Anesthesiology 104:261-266, 2006 8. Campos JH, Hallam EA, Ueda K: Training in placement of the left-sided double-lumen tube among non-thoracic anaesthesiologists: Intubation model simulator versus computer-based digital video disc, a randomised controlled trial. Eur J Anaesthesiol 28:169-174, 2011