Corresponding author:
PD Dr. med. Simon H. Sündermann, MD
Charité- Universitaetsmedizin Berlin
Department of Cardiovascular Surgey
Augustenburger Platz 1
13353 Berlin
Phone: +49 30 450 565841
Email: simon.suendermann@charite.de
Web: https://herzchirurgie.charite.de/
https://tavi.charite.de/
In the study by Kenichi and colleagues published in this issue of the Journal of Cardiac Surgery, a pig’s heart was filled with silicone, subjected to a CT scan and the images imported into special software to display it as a virtual 3D model and measure the relevant anatomical structures. What sounds a little strange at first glance makes perfect sense for a few reasons: Cardiac imaging has become an important part of procedure planning in cardiovascular medicine, especially for structural heart but also coronary artery disease1,2. Image modalities that play a crucial role in this context are cardiac computed tomography (CCT)3 , echocardiography4–6and cardiac magnetic resonance imaging (CMRI)7. Each modality has its own advantages and disadvantages. While CCT has a very high spatial resolution, echocardiography and CMR can be used to perform functional examinations of the heart. To take advantage of different modalities simultaneously, fusion of imaging modalities has been proposed, e.g. fusion of echocardiography and fluoroscopy for mitral interventions8or fusion of invasive coronary angiography and myocardial perfusion imaging9and others. With the increasing computing power of computers, the use of medical images is now reaching new possibilities: The available data is used to segment cardiac structures or even whole hearts10–12. These geometric models can then be used for further applications, such as simulating cardiac interventions13–15or 3D printing to better understand complex anatomies. The latter became popular before interventions involving congenital heart defects16, as these are highly individualised from patient to patient. Using a 3D-printed model, surgical planning can be done ”hands-on”. But the use of 3D models can also be very helpful in adult cardiac surgery, e.g. mitral valve repair, where the anatomy on the arrested heart can be very complex to assess17. In addition to sophisticated modelling of the anatomical structures of the heart and therapy simulation, a third application of the use of high-quality cardiac images can be seen in education and training. Here, too, virtual visualisations of three-dimensional images, but also 3D-printed models can be used. For example, training ventricular septal defect repair on a silicone model of a heart is described as a useful method for training18. In adult cardiac procedures, 3D-printed models can help in planning access, for example in surgical minimally invasive aortic valve replacement19 or to avoid complications in transcatheter aortic valve implantation20. The very recent trend is to integrate 3D models into virtual reality applications, as described in the paper by Kenichi and colleagues presented in this issue. Advantages of such tools are the possibilities to interact with 3D images with little effort. Models can be rotated in all directions, zoomed in and out and measurements can be taken. In addition, 3D models can be used in augmented reality scenarios, e.g. for decision support in complex surgical procedures21.
Despite numerous achievements in this field, there are still some limitations to this advanced technology: First of all, the availability of such tools is limited. Cost can play a role. However, another important factor is the ”unfriendly” user interface of many applications and the complicated integration into the clinical environment. Development is sometimes done without sufficient communication with the clinician, who is the end user. However, this problem arises on both sides of the street. Clinicians often have limited interest in new tools that could help facilitate procedures in the future, but are not yet finally developed. The applications are at an early stage and still have many flaws and drawbacks. Very few are available as ready-to-use tools. Paradoxically, much input from clinicians would be important on the way to a ready-to-use solution, but this is often lacking. Another fact that could lead to mistrust of the new tools is the very small number of patients that are often included in studies to validate the tools. A very important feature of applications used in a medical context is of course the accuracy of the models and measurements. In many studies, the validation of the models is not convincing precisely because of the small number of patients and images that are used to develop and validate the models which may be enough to produce nice pictures and prove the feasibility of certain applications or when the model is used for training purposes. However, as soon as it becomes clinically relevant, the accuracy of the models must be close to one hundred percent. The method described in this paper is an interesting concept for an alternative method to validate the accuracy of a particular tool, in this case virtual reality modelling. The model was compared with other standard measures like 3D-MPR assessment and with manual measurements made possible by using the silicon cast of the original anatomy. Unfortunately, only the feasibility on one model was shown and the measurements were not repeated with multiple hearts or with multiple investigators performing the measurements. However, such a method has the potential to be useful for validation purposes, as the silicon model is naturally more available and consistent than patient data. The development of methods that do not require patient involvement and produce the same results is highly desirable. A big step has already been taken with the advent of machine learning algorithms used in cardiovascular modelling. Several entities are currently being studied, for example, aortic aneurysm. Liang et al. used datasets from twenty-five patients to sample 729 representative aortic shapes from all shape distributions described by a statistical shape model. Rupture risk was calculated using finite element analysis and the backward displacement method22. They could show that their approach is much faster than other methods and had a high risk classification accuracy. Currently, modelling heart structures and simulating therapies is already possible. Many concepts are on the way and developing rapidly. To be convincing and relevant in daily clinical routine, comprehensible validation to prove accuracy is essential. Otherwise, it is dangerous for clinicians to rely on virtual models. The path is already paved, but currently cardiac modelling, simulation and virtual reality are still in development and need to be considered as additional tools for planning and decision-making in cardiac interventions. In the near future, the incredible opportunities presented by such techniques are likely to play a major role in cardiovascular care. Small steps and studies like the one by Kenichi and colleagues are very important to continue on this path, so that virtual reality arrives in medicine and is not reserved for video game lovers.

Literature

1. Di Carli MF, Geva T, Davidoff R. The Future of Cardiovascular Imaging. Circulation2016;133:2640–2661.
2. Edvardsen T, Asch FM, Davidson B, Delgado V, DeMaria A, Dilsizian V, Gaemperli O, Garcia MJ, Kamp O, Lee DC, Neglia D, Neskovic AN, Pellikka PA, Plein S, Sechtem U, Shea E, Sicari R, Villines TC, Lindner JR, Popescu BA. Non-Invasive Imaging in Coronary Syndromes: Recommendations of The European Association of Cardiovascular Imaging and the American Society of Echocardiography, in Collaboration with The American Society of Nuclear Cardiology, Society of Cardiovascular Computed Tomography, and Society for Cardiovascular Magnetic Resonance. J Am Soc Echocardiogr2022;35:329–354.
3. Pontone G, Rossi A, Guglielmo M, Dweck MR, Gaemperli O, Nieman K, Pugliese F, Maurovich-Horvat P, Gimelli A, Cosyns B, Achenbach S. Clinical applications of cardiac computed tomography: a consensus paper of the European Association of Cardiovascular Imaging—part II. Eur Heart J Cardiovasc Imaging 2022;23:e136–e161.
4. Agricola E, Meucci F, Ancona F, Pardo Sanz A, Zamorano JL. Echocardiographic guidance in transcatheter structural cardiac interventions. EuroIntervention2022;17:1205–1226.
5. Barreiro-Perez M, Caneiro-Queija B, Puga L, Gonzalez-Ferreiro R, Alarcon R, Parada JA, Iñiguez-Romo A, Estevez-Loureiro R. Imaging in Transcatheter Mitral Valve Replacement: State-of-Art Review. J Clin Med Res2021;10.
6. Sengupta A, Alexis SL, Zaid S, Tang GHL, Lerakis S, Martin RP. Imaging the mitral valve: a primer for the interventional surgeon. Ann Cardiothorac Surg2021;10:28–42.
7. Ismail TF, Strugnell W, Coletti C, Božić-Iven M, Weingärtner S, Hammernik K, Correia T, Küstner T. Cardiac MR: From Theory to Practice. Front Cardiovasc Med2022;9:826283.
8. Sündermann SH, Biaggi P, Grünenfelder J, Gessat M, Felix C, Bettex D, Falk V, Corti R. Safety and feasibility of novel technology fusing echocardiography and fluoroscopy images during MitraClip interventions. EuroIntervention2014;9:1210–1216.
9. Xu Z, Tang H, Malhotra S, Dong M, Zhao C, Ye Z, Zhou Y, Xu S, Li D, Wang C, Zhou W. Three-dimensional Fusion of Myocardial Perfusion SPECT and Invasive Coronary Angiography Guides Coronary Revascularization. J Nucl Cardiol 2022.
10. Tautz L, Neugebauer M, Hüllebrand M, Vellguth K, Degener F, Sündermann S, Wamala I, Goubergrits L, Kuehne T, Falk V, Hennemuth A. Extraction of open-state mitral valve geometry from CT volumes. Int J Comput Assist Radiol Surg2018;13:1741–1754.
11. Baillargeon B, Rebelo N, Fox DD, Taylor RL, Kuhl E. The Living Heart Project: A robust and integrative simulator for human heart function. Eur J Mech A: Solids 2014;48:38–47.
12. Jafari A, Pszczolkowski E, Krishnamurthy A. A framework for biomechanics simulations using four-chamber cardiac models. J Biomech 2019;91:92–101.
13. Finotello A, Gorla R, Brambilla N, Bedogni F, Auricchio F, Morganti S. Finite element analysis of transcatheter aortic valve implantation: Insights on the modelling of self-expandable devices. J Mech Behav Biomed Mater2021;123:104772.
14. Walczak L, Tautz L, Neugebauer M, Georgii J, Wamala I, Sündermann S, Falk V, Hennemuth A. Interactive editing of virtual chordae tendineae for the simulation of the mitral valve in a decision support system. Int J Comput Assist Radiol Surg 2020.
15. Choi A, Rim Y, Mun JS, Kim H. A novel finite element-based patient-specific mitral valve repair: virtual ring annuloplasty. Biomed Mater Eng2014;24:341–347.
16. Kiraly L, Shah NC, Abdullah O, Al-Ketan O, Rowshan R. Three-Dimensional Virtual and Printed Prototypes in Complex Congenital and Pediatric Cardiac Surgery-A Multidisciplinary Team-Learning Experience. Biomolecules2021;11.
17. Yang Y, Wang H, Song H, Hu X, Hu R, Cao S, Guo J, Zhou Q. A soft functional mitral valve model prepared by three-dimensional printing as an aid for an advanced mitral valve operation. Eur J Cardiothorac Surg2022;61:877–885.
18. Mattus MS, Ralph TB, Keller SMP, Waltz AL, Bramlet MT. Creation of Patient-Specific Silicone Cardiac Models with Applications in Pre-surgical Plans and Hands-on Training. J Vis Exp 2022.
19. Wamala I, Brüning J, Dittmann J, Jerichow S, Weinhold J, Goubergritis L, Hennemuth A, Falk V, Kempfert J. Simulation of a Right Anterior Thoracotomy Access for Aortic Valve Replacement Using a 3D Printed Model. Innovations2019;14:428–435.
20. Yang C, Song G, Niu G, Wu Y. Coronary protection for the small left coronary sinus during transcatheter aortic valve replacement: a case report. Eur Heart J Case Rep 2022;6:ytac011.
21. Chen L, Tang W, John NW. Real-time geometry-aware augmented reality in minimally invasive surgery. Healthc Technol Lett 2017;4:163–167.
22. Liang L, Liu M, Martin C, Elefteriades JA, Sun W. A machine learning approach to investigate the relationship between shape features and numerically predicted risk of ascending aortic aneurysm. Biomech Model Mechanobiol2017;16:1519–1533.