Proceedings of the Institution of Mechanical Engineers, Part H: Journal of Engineering in Medicine http://pih.sagepub.

com/

Injection-Moulded Models of Major and Minor Arteries: The Variability of Model Wall Thickness Owing to Casting Technique
T O'Brien, L Morris, M O'Donnell, M Walsh and T McGloughlin Proceedings of the Institution of Mechanical Engineers, Part H: Journal of Engineering in Medicine 2005 219: 381 DOI: 10.1243/095441105X34347 The online version of this article can be found at: http://pih.sagepub.com/content/219/5/381

Published by:
http://www.sagepublications.com

On behalf of:

Institution of Mechanical Engineers

Additional services and information for Proceedings of the Institution of Mechanical Engineers, Part H: Journal of Engineering in Medicine can be found at: Email Alerts: http://pih.sagepub.com/cgi/alerts Subscriptions: http://pih.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://pih.sagepub.com/content/219/5/381.refs.html

>> Version of Record - May 1, 2005 What is This?

Downloaded from pih.sagepub.com at SWETS WISE ONLINE CONTENT on May 29, 2012

TECHNICAL NOTE

381

Injection-moulded models of major and minor arteries: the variability of model wall thickness owing to casting technique
T OBrien*, L Morris, M ODonnell, M Walsh, and T McGloughlin Centre for Applied Biomedical Engineering Research and Materials and Surface Science Institute, Department of Mechanical and Aeronautical Engineering, University of Limerick, Limerick, Ireland The manuscript was received on 3 September 2004 and was accepted after revision for publication on 19 May 2005. DOI: 10.1243/095441105X34347

Abstract: Cardiovascular disease of major and minor arteries is a common cause of death in Western society. The wall mechanics and haemodynamics within the arteries are considered to be important factors in the disease formation process. This paper is concerned with the development of an ecient computer-integrated technique to manufacture idealized and realistic models of diseased major and minor arteries from radiological images and to address the issue of model wall thickness variability. Variations in wall thickness from the original computer models to the nal castings are quantied using a CCD camera. The results found that wall thickness variation from the major and minor idealized artery models to design specication were insignicant, up to a maximum of 16 per cent. In realistic models, however, dierences were up to 23 per cent in the major arterial models and 58 per cent in the minor arterial models, but the wall thickness variability remained within the limits of previously reported wall thickness results. It is concluded that the described injection moulding procedure yields idealized and realistic castings suitable for use in experimental investigations, with idealized models giving better agreement with design. Wall thickness is variable and should be assessed after the models are manufactured. Keywords: vascular models, injection moulding, wall thickness 1 INTRODUCTION Cardiovascular disease aects both major [1, 2] and minor arteries [3, 4]. It is frequently reported that the haemodynamics within the artery and the wall mechanics of the artery are major factors in the disease formation process [5]. This study is concerned with the development and validation of a CAD/CAM technique that enables concurrent investigation of both numerical and experimental models for both haemodynamic and wall mechanics investigations. A notable shortcoming in many previously reported studies using silicone rubber and wax casting techniques [6, 7] was their failure to address the issue of wall thickness quantication in the modelling process. This study aims to address this.
* Corresponding author: Centre for Applied Biomedical Engineering Research and Materials and Surface Science Institute, Department of Mechanical and Aeronautical Engineering, University of Limerick, Limerick, Ireland. email: thomas.obrien@ ul.ie

The objective of this study was to develop a materially ecient injection moulding technique and to assess the relative dimensional accuracies associated with the modelling of both major and minor models, and idealized and realistic models. While the injection moulding procedure is widely in use today, the dimensional accuracy of the resulting model may vary considerably, depending on the arterial location for which it is used.

2 MATERIALS AND METHODS 2.1 Models of interest Brief descriptions of previously presented computer solid models of the aortic aneurysm [8] and the graft/artery bypass junction [9] follow (Fig. 1). The major dimensions of the idealized abdominal aortic aneurysm model were acquired from the EUROSTAR data registry centre [10], the aorta having a 24 mm lumen diameter, the iliac being 12 mm in diameter,
Proc. IMechE Vol. 219 Part H: J. Engineering in Medicine

H05604 IMechE 2005

Downloaded from pih.sagepub.com at SWETS WISE ONLINE CONTENT on May 29, 2012

382

T OBrien, L Morris, M ODonnell, M Walsh, and T McGloughlin

Fig. 1 Illustration of the idealized and realistic AAA models and the idealized and realistic bypass junction models used in the study

with a 2 mm thick wall throughout the entire model. The realistic abdominal aortic aneurysm model was acquired using a Siemens spiral CT scanner [8] and was designed to have a 2 mm thick wall. The idealized femoral artery bypass junction model was generated from previously published models [11] and has a 6 mm artery and graft lumen diameter with 1 mm thick walls. The realistic femoral artery junction model (Fig. 1) was acquired using an MRI scanner (Imperial College London), and a 1 mm thick wall was specied [12]. There are a number of reasons why the wall thickness should be uniform and controlled. 1. Wall thickness for dierent arteries is generally quoted as a specic value [1317]. In order to create idealized or realistic models with similar pressurestrain elastic responses as quoted in the literature it is necessary to model the walls as being uniform in thickness. 2. In the present study, the two idealized models are developed from computational models that
Proc. IMechE Vol. 219 Part H: J. Engineering in Medicine

assumed uniform wall thickness. Therefore, if validation of the results of these models is required using experimental methods, it is necessary that the wall thickness be uniform and controlled. Previous work [18, 19] presented models of the carotid artery and the coronary artery developed from casts with uniform wall thickness throughout. 3. For idealized models, the wall thickness should be uniform and accurately controlled, as all other dimensions, e.g. lumen diameter, artery curvature, anastomotic angles, etc., are accurately specied. 2.2 Model manufacture The compliant models were cast using silicone rubber in order to produce models that could be investigated in ow measurement experiments using laser Doppler anemometry and particle image velocimetry. In addition, previous work [9] has found that the pressurestrain elastic response of silicone rubber femoral artery models is similar to that reported in the literature at certain physiologically realistic
H05604 IMechE 2005

Downloaded from pih.sagepub.com at SWETS WISE ONLINE CONTENT on May 29, 2012

Injection-moulded models of major and minor arteries

383

pressures. Other authors have also used silicone rubber for creating arterial models [6]. Silicone rubber has a low viscosity, enabling it to be used for casting thin-walled models. Two moulds were manufactured for each model, one for the lumen cast and one for the wall cast. ProEngineer was used to specify the moulds. Each model was subtracted from a workpiece block to which was added necessary supports, guide-pin holes, and vents. The mould block was then split, and each half-mould was exported as a data le to ProManufacture (Parametric Technology Corporation) where the necessary CNC machining parameters were set. The moulds were then machined from aluminium alloy blocks. A casting wax, Texwax FG 890 (Texaco, Germany), was used to make the lumen casts. The lumen casts were reinforced with 1 mm diameter steel wire to prevent sagging of the casts during the injection and moulding phases. The wall mould was coated with mould release formula (Ambersil Formula 8) while the wax lumen cast was coated with releasing agent (Wacker Mould Release). Support for the lumen casts within the wall moulds were provided by design in the ProEngineer part models for the moulds. At the inlets/outlets of each model, an appropriate supporting section was added. The wall moulds were pre-heated to 32 C. The lumen casts were then placed into the moulds, which were then clamped. The silicone rubber (Wacker SilGel 601) was prepared, air bubbles were removed using a vacuum chamber, and it was then injected into the preheated mould. The mould was then transferred to an oven where curing of the silicone rubber took place at 53 C. Once cured, the model was removed from the mould and the wax was melted out at 100 C. The casting was cleaned and dried.

3.2 Model wall thicknesses The wall thickness for each slice was measured at four locations around the edge (equispaced at 90). The results were then averaged for each artery of interest and the standard deviation was calculated as shown in Table 1. The averaged results were then used to determine the dierence between the resulting wall thickness and the thickness that was specied in the design.

4 DISCUSSION The method presented describes a procedure for manufacturing compliant walled, transparent arterial models from corresponding computer solid models for use in experimental investigations of vessel haemodynamics and wall mechanics. Previous studies have used similar techniques to manufacture compliant walled models. However, in many cases, failure to address the issue of wall thickness variance due to wax shrinkage and rubber expansion was evident. In an experimental study of a realistic coronary artery [4], a shadowgraph of the resulting crosssectional areas was taken, but the wall thickness was not measured. A study by Deters et al. [20] used a silicone rubber casting method to model a realistic aortic bifurcation without addressing the wall thickness issue. A carotid bifurcation model [18] chose a wall thickness of 3 mm in a compliant carotid cast but failed to quantify the resulting thickness. In a coronary artery study [19] an estimation of the resulting wall thickness from a casting process was presented. Chong et al. [7] presented a computerintegrated technique to manufacture realistic graft/ artery bypass junctions from radiological images. However, while the study concluded that vessel shape and diameter were preserved, quantication of wall thickness was not mentioned. The results of the present study show that there is a need to assess the resulting wall thickness for silicone rubber casts. In this study, idealized and realistic major arteries were found to have wall thickness dierences from design of up to 23 per cent for the iliac artery as shown in Table 2. Idealized models proved to provide the best agreement with design, from 14 per cent in the large-diameter aorta to 16 per cent in the small-diameter femoral. However, for the realistic models, the thin-walled, minor arteries proved to have a wall thickness signicantly dierent from designin the case of the torturous saphenous vein graft, up to 58 per cent greater than design (Table 2).
Proc. IMechE Vol. 219 Part H: J. Engineering in Medicine

3 RESULTS 3.1 Sectioning of models The models were sectioned with a scalpel at a number of axial locations in order to establish how well their dimensions concurred with those of the computer models. In each model, 1 mm thick slices were taken at an interslice spacing of 4 mm and the wall thickness was measured using a CCD camera (ME-46 Messphysik Laborgerate GmbH). The images were acquired (Fig. 2) and thresholded, and the thickness was established using Scion Image (Scion Corporation). The pixel dimension was established using a 10 mm side calibration cube.
H05604 IMechE 2005

Downloaded from pih.sagepub.com at SWETS WISE ONLINE CONTENT on May 29, 2012

384

T OBrien, L Morris, M ODonnell, M Walsh, and T McGloughlin

Fig. 2 CCD captured images of the slices taken for each of the dierent vessels in the models under study

Previously reported thickness of artery walls vary considerably. Reports of aortic wall thickness range from 1.1 mm [13] to 2 mm [14] and 3.4 mm [15]. Thicknesses of 0.71.4 mm for the coronaries [16],
Proc. IMechE Vol. 219 Part H: J. Engineering in Medicine

0.511.03 mm for the carotid [17], and an intimamedia wall thickness of 0.832.07 mm in the femoral artery [12] have been reported. This study has produced realistic aortas with a wall thickness of
H05604 IMechE 2005

Downloaded from pih.sagepub.com at SWETS WISE ONLINE CONTENT on May 29, 2012

Injection-moulded models of major and minor arteries

385

Table 1 Wall thickness measurements at four locations on the wall slices. The nature of the variability of the wall thicknesses for each slice is evident from the measurement locations
Wall thickness (mm)
Slice 1
Idealized aorta 0.23 mm/pixel Top 2.07 Bottom 2.53 Left 2.53 Right 2.53
Average Standard deviation 2.415 0.20

Slice 2
2.3 2.3 2.53 2.53
2.415 0.12 3.15 1.89 2.73 2.31 2.52 0.47

Slice 3
2.3 2.53 2.3 2.07
2.3 0.16 2.73 1.89 2.31 1.89 2.205 0.35
2.38 2.52 2.94 3.08 2.73 0.29 2.4 2.4 2.4 2.25 2.3625 0.06 1.6 0.8 1.04 1.28 1.18 0.30 1.52 1.76 1.76 1.28 1.58 0.20

Slice 4
2.07 2.3 2.3 2.3
2.2425 0.10 2.52 2.1 2.31 2.52 2.3625 0.17
1.82 2.8 2.8 1.12 2.135 0.71 2.1 2.55 2.4 1.65 2.175 0.34 1.6 0.72 0.96 1.12 1.1 0.32 1.52 1.6 1.36 1.44 1.48 0.09 1.9 1.7 1.5 1.2 1.575 0.26

Slice 5
2.07 2.07 2.07 1.84
2.0125 0.10 2.73 3.15 2.1 2.73 2.6775 0.37
3.36 3.36 2.52 0.84 2.52 1.03 2.1 2.25 2.25 2.4 2.25 0.11 1.44 0.88 0.96 1.28 1.14 0.23 1.28 1.76 1.28 1.6 1.48 0.21 1.9 1.3 1.3 1.4 1.475 0.25

Table 2 Percentage dierence of the wall thickness measurements in the resulting casts from that specied in the original computer model. All models experience an increase in the thickness of the wall over the original specication owing to expansion of the silicone rubber on curing, and also owing to contraction of the wax
Lumen diameter (mm)
Ideal models Aorta Iliac Femoral Realistic models Aorta Illiac Saphenous Femoral 24 12 6 18 10 10 6

Design wall thickness (mm)

2 2 1 2 2 1 1

% dierence in wall thickness to design

14 14 16 20 23 58 53

Realistic aorta 0.21 mm/pixel Top 2.52 Bottom 1.89 Left 2.31 Right 2.1 Average Standard deviation 2.205 0.23

Realistic right iliac 0.14 mm/pixel Top 2.38 2.94 Bottom 2.24 2.8 Left 2.52 2.38 Right 1.68 2.52 Average Standard deviation 2.205 0.32 2.66 0.22

Idealized right iliac 0.15 mm/pixel Top 1.5 2.7 Bottom 2.4 2.4 Left 2.25 2.7 Right 2.1 2.4 Average Standard deviation 2.0625 0.34 2.55 0.15

Idealized femoral artery 0.08 mm/pixel Top 1.28 1.36 Bottom 1.2 0.96 Left 0.88 1.12 Right 1.52 1.28 Average Standard deviation 1.22 0.23 1.18 0.15

Realistic femoral artery 0.08 mm/pixel Top 1.52 1.36 Bottom 1.36 2 Left 1.2 1.52 Right 1.68 1.76 Average Standard deviation 1.44 0.18 1.66 0.24

Realistic saphenous vein graft 0.1 mm/pixel Top 1.6 1.9 1.5 Bottom 1.7 1.9 1.7 Left 1.5 1.6 1.6 Right 1.4 1.3 1.7 Average Standard deviation 1.55 0.11 1.675 0.25 1.625 0.08

resulting wall thickness measurements do lie within previously reported ranges. A study by Loth et al. [21] used a scaled-up cast of a vascular graft model to increase the spatial resolution of the ow measurement technique, which is also an option available to researchers to increase the accuracy of the modelling process. A number of factors have been identied that cause variability in wall thickness. The most signicant is the expansion of the silicone rubber upon curing. In all cases, the wall thickness was found to be greater than that of the design specication. In addition, the wax casts were noted to shrink slightly upon solidication. Finally, it should be noted that renements to the technique may reduce the dierence of realistic model thickness from those specied in the design. The dierences are due to the highly tortuous nature of the realistic models and are to be expected when compared with the results of the more easily manufactured idealized models. Nevertheless, the technique is capable of producing models suitable for use in experimental investigations, although wall variance must be assessed in the nal models.

5 CONCLUSION Compliant idealized and realistic models of various arterial geometries were manufactured using a dened CAM injection moulding technique. It was found that there was a degree of variability in the wall thickness reproducibility of the modelling procedure. While the wall thicknesses lay within
Proc. IMechE Vol. 219 Part H: J. Engineering in Medicine

2.390.32 mm, iliac arteries of 2.450.51 mm, idealized femoral arteries of 1.160.25 mm, and realistic femoral arteries of 1.530.18 mm. While signicant dierences exist between design and result, the
H05604 IMechE 2005

Downloaded from pih.sagepub.com at SWETS WISE ONLINE CONTENT on May 29, 2012

386

T OBrien, L Morris, M ODonnell, M Walsh, and T McGloughlin

previously reported limits, it was found that the idealized models gave best reproduction of the corresponding computer models. The diculty associated with producing realistic compliant models of diseased arteries was evident, in particular with modelling minor arteries. The wall thickness variability of injection-moulded casts of major and minor arteries should be assessed following the casting process.

10

ACKNOWLEDGEMENTS
11

The authors would like to thank the Department of Radiology, Mid-Western Regional Hospital, Limerick, Ireland, and Dr S. J. Sherwin of the Department of Aeronautics, Imperial College, London.
12

REFERENCES
1 Finol, E. A. and Amon, C. H. Blood ow in abdominal aortic aneurysms: pulsatile ow hemodynamics. J. Biomech. Engng, 2001, 123, 474484. 2 Morris, L., Delassus, P., Walsh, M., and McGloughlin, T. A mathematical model to predict the in vivo pulsatile drag forces acting on bifurcated stent grafts used in endovascular treatment of abdominal aortic aneurysms (AAA). J. Biomechanics, 2004, 37, 1087 1095. 3 Walsh, M., McGloughlin, T., Liepsch, D. W., OBrien, T., Morris, L., and Ansari, A.R. On using experimentally estimated wall shear stresses to validate numerically predicted results. Proc. Instn Mech. Engrs, Part H: J. Engineering in Medicine, 2003, 217, 7790. 4 Cho, Y. I., Back, L. H., Crawford, D. W., and Cuel, R. F. Experimental study of pulsatile and steady ow through a smooth tube and an atherosclerotic coronary artery casting of man. J. Biomechanics, 1983, 16(11), 933946. 5 Nerem, R. M. Vascular uid mechanics, the arterial wall and atherosclerosis. J. Biomech. Engng, 1992, 114, 274281. 6 Liepsch, D., Moravec, S. T., and Baumgart, R. Some ow visualisation and LDV measurements in a trueto-scale elastic model of a human aortic arch a new model technique. Biorheology, 1992, 29, 563580. 7 Chong, C. K., Rowe, C. S., Sivanesan, S., Rattray, A., Black, R. A., Shortland, A. P., and How, T. V. Computer aided design and fabrication of models for in vitro studies of vascular uid dynamics. Proc. Instn Mech. Engrs, Part H: J. Engineering in Medicine, 1999, 213, 14. 8 Morris, L., Delassus, P., Grace, P., Wallis, F., Walsh, M., and McGloughlin, T. Eects of at, parabolic, and realistic steady ow inlet proles on idealised and realistic stent graft ts through

13 14

15

16

17

18

19 20

21

Abdominal Aortic Aneurysms (AAA), Med. Eng. Phys. (in press). OBrien, T., Walsh, M. and McGloughlin, T. On reducing abnormal hemodynamics in the femoral end-to-side anastomosis: the inuence of mechanical factors. Ann. Biomed. Eng., 2005, 33(3), 310322. Laheij, R. C., Van Marrewijk, C., and Buth, J. Progress report on the procedural and follow-up results of 3413 patients who received stent graft treatment for infra-renal abdominal aortic aneurysms for a period of 6 years. EUROSTAR Data Registry Centre Publication, 2001. Walsh, M. T., Kavanagh, E. G., OBrien, T., Grace, P. A. and McGloughlin, T. On the existence of an optimum end-to-side junctional geometry in peripheral bypass surgery a computer generated study. Eur. J. Vascular and Endovascular Surg., 2003, 26, 649656. Giordana, S., Sherwin, S. J., Peiro, J., Doorly, D. J., Papaharilaou, Y., Caro, C. G., Watkins, N., Cheshire, N., Jackson, M., Bicknall, C., and Zervas, V. Automated classication of peripheral distal by-pass geometries reconstructed from medical data. J. Biomechanics, 2005, 38(1), 4762. Nichols, W. W. and O Rourke, M. F. McDonalds Blood Flow in Arteries, 4th edition, 1998 (Arnold). Raghavan, M. L., Vorp, D. A., Federle, M. P., Makaroun, S., and Webster, M. W. Wall stress distribution on three-dimensionally reconstructed models of human abdominal aortic aneurysm. J. Vascular Surg., 2000, 31, 760769. Gullace, G. and Rua, F. Eect of Pravastatin on abdominal aorta and carotid wall thickness in dyslipidemia. J. Am. Coll. Cardiology, 1995, 25(2), 369A370A. Ozolanta, I., Tetere, G., Purinya, B., and Kasyanov, V. Changes in the mechanical properties, biochemical contents and wall structure of the human coronary arteries with age and sex. Med. Engng and Phys., 1998, 20, 523533. Baron, Y. M., Brincat, M., and Galea, R. Carotid artery wall thickness in women treated with hormone replacement therapy. Maturitas, 1997, 27, 4753. Anayiotos, A. S., Jones, S. A., Giddens, D. P., Glagov, S., and Zarins, C. K. Shear stress at a compliant model of the human carotid bifurcation. J. Biomech. Engng, 1994, 116, 98106. Weston, M. W. and Tarbell, J. M. Wall shear rate measurements in an elastic curved artery model. Biorheology, 1997, 34(1), 117. Deters, O. J., Bargeron, C. B., Mark, F. F., and Friedman, M. H. Measurement of wall motion and wall shear in a compliant arterial cast. J. Biomech. Engng, 1986, 108, 355358. Loth, F., Jones, S. A., Giddens, D. P., Bassiouny, H. S., Glagov, S., and Zarins, C. K. Measurements of velocity and wall shear stress inside a PTFE vascular graft model under steady ow conditions. J. Biomech. Engng, 1997, 119, 187194.

Proc. IMechE Vol. 219 Part H: J. Engineering in Medicine

Downloaded from pih.sagepub.com at SWETS WISE ONLINE CONTENT on May 29, 2012

H05604 IMechE 2005