3D Printing and 3D Virtual Models for Surgical and Percutaneous

Planning of Congenital Heart Diseases

Katia Capellini

1,2

, Paolo Tripicchio

5 a

, Emanuele Vignali

1,2

, Emanuele Gasparotti

1,2

Lamia Ait Ali

3 b

, Massimiliano Cantinotti

4 c

, Duccio Federici

, Giuseppe Santoro

Francesca Alfonzetti

, Chiara Evangelista

, Camilla Tanca

and Simona Celi

1 d

BioCardioLab, Bioengineering Unit, Fondazione Toscana ”G. Monasterio”, Massa, Italy

Department of Information Engineering, University of Pisa, Pisa, italy

Institute of Clinical Physiology, CNR-Regione Toscana, Massa, Italy

Paediatric Cardiology Unit, Fondazione Toscana ”G. Monasterio”, Massa, Italy

Perceptual Robotics Lab, TeCIP Institute, Scuola Superiore Sant’Anna, Pisa, Italy

Keywords:

Virtual Model, Virtual Reality, 3D Printing, Complex Congenital Heart Diseases.

Abstract:

Despite increasing evidence of their utility, 3D models have never been extensively tested so far in pediatric

cardiac surgery planning. 3D models may offer advantages over traditional imaging examinations: 1) a deeper

understanding of 3D anatomy in complex defects allowing visual and tactile inspection from any point of view,

2) the possibility to interact with a tangible replica of the real organs, 3) the surgical planning and simulation

maneuvers on the printed and virtual model, and 4) interaction with anatomical structures thank to Virtual

Reality technologies. The work aims to test and compare the accuracy and the incremental diagnostic value of

3D printed and virtual models in patients undergoing cardiac surgery for CHDs.

1 INTRODUCTION

In the last years, the interest in 3D printed models

is increased in numerous medical ﬁelds (Vukicevic

et al., 2017) both for the operative planning of dif-

ferent surgical approaches and the development of

custom devices based on the patient-speciﬁc cases

(Kurenov et al., 2015)(Sun et al., ). Congenital heart

diseases (CHDs) are an ideal ﬁeld to test the poten-

tialities, accuracy, reproducibility and clinical effec-

tiveness of 3D technologies due to the complexity

and diversity of cases, the need for a complete repre-

sentation of intra/extra-cardiac anatomy, and of per-

sonalized interventional approaches and size materi-

als (Cantinotti et al., 2017). In complex CHDs the

understanding of the 3D spatial relationship in an un-

usual anatomical arrangement is certainly a major dif-

ﬁculty. Currently medical imaging is able to provide

https://orcid.org/0000-0003-3225-2782

https://orcid.org/0000-0003-1672-5308

https://orcid.org/0000-0002-4671-9606

https://orcid.org/0000-0002-7832-0122

functional and anatomical detailes with hight resolu-

tion and accuracy (Burchill et al., 2017) (Greil et al.,

2017) (Celi et al., 2017) and are used as starting point

for several advanced studies based on integration with

numerical models (Celi and Berti, 2013) (Celi et al.,

2013) (Capellini et al., 2018). Despite all these ad-

vances in the research ﬁeld, currently, for the rou-

tinely diagnosis of CHDs there is a strong need to in-

troduce interactive tools in clinical practise. In fact,

all current 3D imaging modalities are not interactive

and don’t allow to manipulate the 3D image or the

projected images. The standard volume rendering ap-

plications included in the image processing worksta-

tions are not able to provide tangible surfaces and

edges; it refers to a technique for generating a visual

representation of data that is contained in a three di-

mensional space. Even if volume rendering is an im-

portant graphics and visualization technique and sev-

eral techniques and algorithms have been developed

to provide high quality visualization (El Seoud and

Mady, 2019), this approach is not able to produce

a mathematical model useful for additional clinical

evaluation and planning. Reconstructed 3D models

Capellini, K., Tripicchio, P., Vignali, E., Gasparotti, E., Ali, L., Cantinotti, M., Federici, D., Santoro, G., Alfonzetti, F., Evangelista, C., Tanca, C. and Celi, S.

3D Printing and 3D Virtual Models for Surgical and Percutaneous Planning of Congenital Heart Diseases.

DOI: 10.5220/0009160702810287

In Proceedings of the 15th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2020) - Volume 3: IVAPP, pages

281-287

ISBN: 978-989-758-402-2; ISSN: 2184-4321

281

can reproduce anatomical details with high accuracy

and permit both virtual and physical manipulation of

the model offering the advantage to simulate surgical

interventions in a real physical environment in terms

of spatial relationship with an adjunct tactile sense.

The key problems in complex CHDs are the small

and often the non conventional dimensions and spa-

tial relationships (Triedman and Newburger, 2016)

(Stout et al., 2019). With this in mind, in addition

to the physician’s knowledge, and his/her experience,

an important role is played by the spatial intelligence,

deﬁned as “a capacity for mentally generating, ro-

tating, and transforming visual image” (Park et al.,

2010). This kind of intelligence is crucial to the ef-

fectiveness of CHD physicians (Gardner, 2008) and

is considered a fundamental element in medical ed-

ucation (Hegarty, 2014) and more in particular for

anatomy education of the cardiovascular system (Ku-

malasari et al., 2017). In this context, research results

(Sajid et al., 1990) have shown the beneﬁts of three-

dimensional approaches for learning. Indeed, physi-

cal interaction with a 3D model allows to understand

the physical structures of the organs and obtain fa-

miliarity with them (Cooper and Taqueti, 2008). In

the study presented by (Maresky et al., 2018) stu-

dents who were exposed to VR demonstrated sig-

niﬁcant improvement in their understanding of car-

diac anatomy. The learning advantage of VR tech-

nology has been proven useful also in the training of

minimally invasive surgical procedures (Konietschke

et al., 2010). Given its complex three-dimensional

structures, cardiac anatomy may be challenging to

grasp. In this context, VR technologies offer immer-

sive and intuitive experiences that allow appreciating

the size differences of such structures and at the same

time to contextualize their relationships. Several VR

applications in cardiovascular medicine education are

currently being explored, for instance, the one de-

veloped in the Stanford Virtual Heart Project (2017)

where such technology is used in the context of pedi-

atric patients’ parents’ education to allow them visu-

alizing their child’s congenital heart disease. These

kinds of applications contributed to the research in

cardiovascular intervention by assisting the physi-

cians in learning and interpretation of cardiovascular

anatomy and pathology, increasing the precision and

reducing the invasiveness of the interventions (Silva

et al., 2018). 3D models may provide to pediatric sur-

geon/interventionalist a direct visualization of com-

plex intra-extra-cardiac anatomy. Moreover, the com-

bined use of 3D virtual and printed models may help

to plan more accurate surgical/interventional strate-

gies and to choose materials of proper size (i.e. con-

duit, ballon, prosthesis, etc) (Moore et al., 2018). Im-

mersive bimanual exploration of virtual models will

complement the understanding and planning capabil-

ities of the printed model, without requiring complex

haptic devices. In a recent study, some advantages

of VR technique respect to the 3D printing approach

have been depicted (Ong et al., 2018), however, ac-

cording to our knowledge, there are no studies in the

literature that investigate the effectiveness of these

two approaches on the same populations of cases.

Furthermore, virtual models, combining immersive

visual and vibrotactile feedback, have been scarcely

tested for pediatric cardiac surgery (Izard et al., 2018),

and no speciﬁc comparison between the 3D virtual

model and 3D printed anatomies have been proposed.

This work aims to perform a comparison of the adop-

tion of 3D printed and 3D virtual models for complex

CHDs defects in terms of the satisfaction of physi-

cians. To extend our investigation, both surgical and

endovascular planning have been investigated.

2 MATERIALS AND METHODS

The workﬂow of this study is reported in Figure 1.

It starts from the segmentation of clinical images to

create the 3D model of cardiac structures affected by

CHDs. The 3D models are printed with different 3D

printing techniques and materials and used as virtual

models in a virtual environment and a VR platform.

Depending on the case, the clinical team investigates

the most suitable surgical or catheter-based procedure

on both the printed and virtual models. This workﬂow

is illustrated for two CHDs cases: aortic coarctation

(CoA) and heart with a complex CHD.

Figure 1: Diagram of the procedural phases of the presented

approach.

IVAPP 2020 - 11th International Conference on Information Visualization Theory and Applications

282

2.1 Image Processing

Twelve Magnetic Resonance (MR) and eight Com-

puted Tomography (CT) volumetric datasets of CHDs

patients scheduled for surgical repair were analyzed.

In order to obtain a 3D model of cardiac structures of

interest, segmentation was performed by adopting dif-

ferent techniques. Semi-automatic segmentation al-

gorithms such us threshold, active contours, and re-

gion growing algorithms were adopted when possible,

together with manual segmentation slice by slice for

more complex regions of interest.

After the segmentation phase, a process of mesh

reﬁnement was necessary to obtain the ﬁnal model

without imperfections and a constant thickness was

assigned. The resulting 3D shell models were both

printed and used in the virtual environment. Figure

2 depicts, as example, the three main phases of this

process in case of a MR dataset: the medical volume

rendering (Figure 2a), the segmentation process (Fig-

ure 2b) and the ﬁnal 3D result (Figure 2c).

2.2 3D Printing

In this study, two different 3D printing strategies

were tested: the fused deposition modeling technique

(FDM) and the stereolithography technique (SLA). A

thermoplastic polyurethane material (TPU Elasto85)

with shore A85 (Standard, 2005) was used in the

FDM approach. A hydrosoluble support polymer

(SSU04) was used due to the presence of cavities in

the models. The models were printed with a layer

thickness of 0.25 mm on a 3ntr A4v4 3D printer. For

the SLA technique, a Form2 3D printer was adopted

with a clear elastic resin with shore A50 (Standard,

2005) and a layer thickness of 0.1 mm. The internal

and external supports were manually removed.

2.3 Virtual Technique

Given the fact that there is no necessity for the data

to be produced in real-time, the virtual techniques in-

volve the production of 3D models and textures that

could take several hours and days. One of the objec-

tives of the VR visualization presented here is instead

that of producing a textured 3D model that is possi-

ble to visualize as soon as possible thus introducing

a trade-off between the accuracy of the representation

and immediate availability of the model. For this rea-

son, the 3D model segmented from MR and CT scans

are passed to a pre-processing step where the mod-

els are optimized for visualization and the texture co-

ordinates for each vertex are generated. A prelimi-

nary texture, not reﬂecting the real organic texture but

with realistic rendering, is produced and applied on

the model on the ﬂy. This allows us to reduce the

timing for the VR simulation to be used, and this is

extremely important giving the time at disposal for

pre-operation analysis. The VR interaction with the

generated 3D model will give the surgeons some tools

to manipulate the 3D scene. In particular, the user

can rotate the view to analyze the 3D structure at its

best. A virtual cutting feature has been implemented

to enable performing some cuts on the 3D model sur-

face to explore the internal cavities and plan possi-

ble surgical exploration of the models. A second im-

plemented tool allows using simple 3D shapes (like

cubes and sphere) to perform a real-time clipping on

the 3D model thus showing internal elements in a non

disruptive way. Examples of these two types of oper-

ation are shown in ﬁgure 3 and 4. The Unity engine

was used to develop both a desktop VR application

and an immersive VR visualization for the Oculus rift

Head Mounted Display (HMD).

2.4 Pre-operative Planning

Every case under exam has been discussed by a mul-

tidisciplinary team (cardiac surgeons, pediatric cardi-

ologists, anesthetists) in two different steps: ﬁrstly,

based on conventional imaging and, secondly, with

the support of the 3D printed and virtual models.

The planning strategies were compared and the added

value of the two additional methods was evaluated.

For this study, the same team has evaluated all the 3D

printed and virtual models.

3 RESULTS

The image segmentation was feasible for all datasets

and the corresponding 3D shell models were gener-

ated for all patients. For the subjects with a CT dataset

an automatic segmentation was practicable due to the

high spatial resolution. The segmentation of the MR

datasets was more complex to be performed. It is

worth to point out that the complexity of this process

was due to the presence of breath and motion arti-

facts and, in general, due to the absence of the contrast

medium.

Case of Coarctation - In Figure 5 an example of

planning for a percutaneous procedure for an artery

affected by CoA is reported. Starting from an MR

dataset (Figure 5a), the model was 3D printed with the

SLA technique and used by the clinician to simulate

the endovascular procedure during the pre-planning

phase.

3D Printing and 3D Virtual Models for Surgical and Percutaneous Planning of Congenital Heart Diseases

283

Figure 2: An example of standard volume rendering (a), images segmentation (b) and the ﬁnal 3D model (c).

Figure 3: Rendering in VR of a virtual heart model with a

custom texture (a) and example of cut applied on the mesh

enabling the display of internal structures (b).

Figure 4: Virtual 3D model (a) and example of model ma-

nipulated through the use of a clipping plane (b).

Due to the complexity of this CHD, two differ-

ent devices were required (Figure 5b): one to occlude

a pseudo-aneurysm at the aortic arch level and one

to recover the lumen of the CoA. Figure 5c depicts

the intra-operative image according to the planning

procedure. In this case the 3D virtual model (Figure

5d) was used to take the measurements to deﬁne the

proper size of devices.

Figure 5: Case report of CoA from images to intervention:

MR image (a), simulation of endovascular treatment on 3D

printed model (b), angiographic image acquired during pro-

cedure (c) and 3D virtual model (d).

Case of Complex CHD - In Figure 6 the planning for

a heart CHD is reported. The surgeon tested the se-

lected plan on the 3D printed model by cutting in the

deﬁned region from the same view that he would have

in the operating room and he explored the heart inside

IVAPP 2020 - 11th International Conference on Information Visualization Theory and Applications

284

(Figure 6a-b) to better understand which parts could

be seen and reached thank to the possibility to inter-

act with the real patient dimensions. Also the possible

surgical strategies were investigated by implementing

a virtual planning (Figure 6c).

Regarding the 3D printing techniques, it has al-

ways been possible to obtain 3D models with high

accuracy by using the FDM approach. This technique

permitted to have 3D deformable heart models to sim-

ulate the operative incisions by the surgeons as re-

ported in Figure 6.

Figure 6: Surgical planninig on printed model (a-b) and on

3D virtual model (c).

The models printed with SLA were required in

case of device insertion and expansion (Figure 5).

In Figure 7 an example of VR application is re-

ported. A user can visualize the organic tissues in 3D

within an HMD, thus exploring the model in the ﬁrst

person. This enables changing the viewpoint in a nat-

ural manner reproducing an immersive experience for

the practitioner. Special tools enable the user to visu-

alize both the surface properties and the internal cav-

ities, thanks to the virtual culling of the 3D mesh.

The preliminary results of clinicians’ feedback

showed that the use of the 3D model in a virtual en-

vironment is the better choice to investigate the eli-

gible CHDs repair strategies due to the possibility to

investigate more sites of access for the operation by

reversibly editing the model. The 3D printed models

allowed the in vitro surgical and percutaneous plan-

ning simulations in time with clinical timing.

4 CONCLUSIONS

Despite recent advances in current imaging tech-

niques for the diagnosis and management of CHDs,

several limitations in 3D visualization remain. The

3D printing technique and the deployment of a virtual

environment for the CHDs models improve the surgi-

cal and percutaneous planning for the pathology treat-

ment. 3D virtual models have proven to be a useful in-

strument to assist the clinical team and the surgeon in

the decision-making procedure for the patient-speciﬁc

best intervention strategy in the case of surgical ap-

proach.

The 3D printed model is used to test the feasibility

of the determined surgical technique by giving clini-

cians the appropriate level of conﬁdence, also related

to the real dimensions of patient anatomic structures,

not reached with biomedical images visualization and

3D virtual models (Batteux et al., 2019).

Regarding the time costs, the segmentation pro-

cess affected all approaches and ranging from less

than an hour for CT datasets with contrast medium to

several hours for some MR datasets. The 3D virtual

simple model is available from the 3D shell model

without any additional times except for the texture

assignment, if required. The VR implementation al-

lowed us to navigate in an immersive context with

an intuitive manipulation of the 3D model. The in-

clusion of 3D model in a virtual reality environment

made possible the model rotation in the 3D space,

measurements on the model to get the real distance

and relationships between the different regions of the

heart or arterial models, the variation of model di-

mensions and the option to cut a speciﬁc portion of

surface model to analyze internal cavities in a non-

destructive manner. This last aspect is particularly

important due to several surgical cuts can be tested

on the same model. Even if a simple texture has been

adopted, this approach reveals an increase of conﬁ-

dence from the clinical team with respect to a single

color model. This same comment has been reported

by clinician comparing the VR model with respect to

the 3D printed one. In fact, it is worth to point out that

the 3D printed model were in a single color with a low

range of color available from the 3D printer manufac-

3D Printing and 3D Virtual Models for Surgical and Percutaneous Planning of Congenital Heart Diseases

285

Figure 7: Example of VR application.

turer. Regarding the 3D printing approach, the time

cost for the 3D printing technique included the print-

ing time and the supports removal time and increases

with the model size and complexity ranging from ten

to twenty-ﬁve hours. The SLA technique presented

the main drawback of internal supports removal that

it resulted in impossible in the cases of the whole

heart model. The supports removal was feasible for

the artery models and in these cases, a simulation of

endovascular procedure for the CHDs treatment was

easily performed. 3D printed models with FDM was

more suitable for a surgical procedure, while the use

of the elastic resin turned out to be the most suitable

to simulate endovascular procedures involving device

expansion.

Moreover, the VR platform and 3D printing mod-

els seemed suitable for medical students’ education

thanks to the possibility to navigate inside the model

and better visualize CHDs (Figure 7). With speciﬁc

attention to the VR environment, usability assessment

of different interaction metaphors is the focus of fu-

ture work, testing different VR setups including a

classical monitor visualization, an immersive simula-

tion trough an HMD and the use of a tablet for navi-

gation and display of 3D virtual content.

ACKNOWLEDGEMENTS

This work is supported by the ”3D VIRTUAL BABY

HEART” project (2018-2020), founded by the Ital-

ian Ministry of Health (grant number: GR-2016-

02365072).

REFERENCES

Batteux, C., Haidar, M., and Bonnet, D. (2019). 3d-printed

models for surgical planning in complex congenital

heart diseases: a systematic review. Frontiers in pe-

diatrics, 7:23.

Burchill, L. J., Huang, J., Tretter, J. T., Khan, A. M., Crean,

A. M., Veldtman, G. R., Kaul, S., and Broberg, C. S.

(2017). Noninvasive imaging in adult congenital heart

disease. Circulation Research, 120(6):995–1014.

Cantinotti, M., Valverde, I., and Kutty, S. (2017). Three-

dimensional printed models in congenital heart dis-

ease. The international journal of cardiovascular

imaging, 33(1):137–144.

Capellini, K., Vignali, E., Costa, E., Gasparotti, E., Bian-

colini, M. E., Landini, L., Positano, V., and Celi, S.

(2018). Computational ﬂuid dynamic study for ataa

hemodynamics: An integrated image-based and radial

basis functions mesh morphing approach. Journal of

Biomechanical Engineering, 140(11).

Celi, S. and Berti, S. (2013). Three-dimensional sensitivity

assessment of thoracic aortic aneurysm wall stress: a

probabilistic ﬁnite-element study. European Journal

of Cardio-Thoracic Surgery, 45(3):467–475.

Celi, S., Martini, N., Pastormerlo, L. E., Positano, V., and

Berti, S. (2017). Multimodality imaging for interven-

tional cardiology. Current Pharmaceutical Design,

23(22):3285–3300.

Celi, S., Vaghetti, M., Palmieri, C., and Berti, S. (2013).

Superﬁcial coronary calcium analysis by oct: Look-

ing forward an imaging algorithm for an automatic 3d

quantiﬁcation. International Journal of Cardiology,

168(3):2958–2960.

Cooper, J. and Taqueti, V. (2008). A brief history of the

development of mannequin simulators for clinical ed-

ucation and training. Postgraduate medical journal,

84(997):563–570.

IVAPP 2020 - 11th International Conference on Information Visualization Theory and Applications

286

El Seoud, M. S. A. and Mady, A. S. (2019). A compre-

hensive review on volume rendering techniques. In

Proceedings of the 2019 8th International Conference

on Software and Information Engineering, ICSIE ’19,

page 126–131, New York, NY, USA. Association for

Computing Machinery.

Gardner, H. E. (2008). Multiple intelligences: New horizons

in theory and practice. Basic books.

Greil, G., Tandon, A. A., Silva Vieira, M., and Hussain, T.

(2017). Noninvasive imaging in adult congenital heart

disease. Frontiers in Pediatrics, 5(6):36.

Hegarty, M. (2014). Spatial thinking in undergraduate sci-

ence education. Spatial Cognition & Computation,

14(2):142–167.

Izard, S. G., Juanes, J. A., Pe

nalvo, F. J. G., Estella, J. M. G.,

Ledesma, M. J. S., and Ruisoto, P. (2018). Virtual re-

ality as an educational and training tool for medicine.

Journal of medical systems, 42(3):50.

Konietschke, R., Tobergte, A., Preusche, C., Tripicchio, P.,

Ruffaldi, E., Webel, S., and Bockholt, U. (2010). A

multimodal training platform for minimally invasive

robotic surgery. In 19th International Symposium in

Robot and Human Interactive Communication, pages

422–427. IEEE.

Kumalasari, L., Yusuf Hilmi, A., and Priyandoko, D.

(2017). The application of multiple intelligence ap-

proach to the learning of human circulatory system.

In Journal of Physics Conference Series, volume 909.

Kurenov, S. N., Ionita, C., Sammons, D., and Demmy,

T. L. (2015). Three-dimensional printing to facilitate

anatomic study, device development, simulation, and

planning in thoracic surgery. The Journal of thoracic

and cardiovascular surgery, 149(4):973–979.

Maresky, H., Oikonomou, A., Ali, I., Ditkofsky, N.,

Pakkal, M., and Ballyk, B. (2018). Virtual real-

ity and cardiac anatomy: Exploring immersive three-

dimensional cardiac imaging, a pilot study in un-

dergraduate medical anatomy education. Clinical

Anatomy, 32.

Moore, R. A., Riggs, K. W., Kourtidou, S., Schneider,

K., Szugye, N., Troja, W., D’Souza, G., Rattan, M.,

Bryant III, R., Taylor, M. D., et al. (2018). Three-

dimensional printing and virtual surgery for congeni-

tal heart procedural planning. Birth defects research,

110(13):1082–1090.

Ong, C. S., Krishnan, A., Huang, C. Y., Spevak, P., Vricella,

L., Hibino, N., Garcia, J. R., and Gaur, L. (2018). Role

of virtual reality in congenital heart disease. Congen-

ital heart disease, 13(3):357–361.

Park, G., Lubinski, D., and Benbow, C. P. (2010). Recog-

nizing spatial intelligence. Scientiﬁc American.

Sajid, A., Ewy, G., Felner, J., Gessner, I., Gordon, M.,

Mayer, J., Shub, C., and Waugh, R. (1990). Cardiol-

ogy patient simulator and computer-assisted instruc-

tion technologies in bedside teaching. Medical educa-

tion, 24(6):512–517.

Silva, J. N., Southworth, M., Raptis, C., and Silva, J. (2018).

Emerging applications of virtual reality in cardiovas-

cular medicine. JACC: Basic to Translational Science,

3(3):420–430.

Standard, A. (2005). D2240,”standard test method for rub-

ber property—durometer hardness”. ASTM Interna-

tional, West Conshohoken, PA, USA.

Stout, K. K., Daniels, C. J., Aboulhosn, J. A., Bozkurt, B.,

Broberg, C. S., Colman, J. M., Crumb, S. R., Dearani,

J. A., Fuller, S., Gurvitz, M., Khairy, P., Landzberg,

M. J., Saidi, A., Valente, A. M., and Hare, G. F. V.

(2019). 2018 aha/acc guideline for the management

of adults with congenital heart disease: A report of the

american college of cardiology/american heart associ-

ation task force on clinical practice guidelines. Circu-

lation, 139(14):e698–e800.

Sun, Z., Lau, I., Wong, Y. H., and Yeong, C. H. Personalized

three-dimensional printed models in congenital heart

disease. Journal of Clinical Medicine, 8(522).

Triedman, J. K. and Newburger, J. W. (2016). Trends in

congenital heart disease. Circulation, 133(25):2716–

2733.

Vukicevic, M., Mosadegh, B., Min, J. K., and Little, S. H.

(2017). Cardiac 3d printing and its future directions.

JACC: Cardiovascular Imaging, 10(2):171–184.

3D Printing and 3D Virtual Models for Surgical and Percutaneous Planning of Congenital Heart Diseases

287