UDK 577.1:616.314-76/-77	ISSN 1580-2949
Original scientific article/Izvirni znanstveni članek	MTAEC9, 46(2)123(2012)
AN RE/RM APPROACH TO THE DESIGN AND MANUFACTURE OF REMOVABLE PARTIAL DENTURES WITH A BIOCOMPATIBILITY ANALYSIS OF THE F75
Co-Cr SLM ALLOY
RE/RM-PRIBLIŽEK, NAČRTOVANJE IN IZDELAVA SNEMLJIVIH DELOV ZOBOVJA Z ANALIZO BIOKOMPATIBILNOSTI ZLITINE
F75 Co-Cr SLM
Danimir P. Jevremovic1, Tatjana M. Puškar2, Igor Budak3, Djordje Vukelic3, Vesna Kojic4, Dominic Eggbeer5, Robert J. Williams5
1Clinic for Prosthodontics, School of Dentistry, Pančevo, University Business Academy, Novi Sad, Serbia 2Clinic for Prosthodontics, Medical Faculty - Department of dentistry, University of Novi Sad, Serbia 3Faculty of Technical Sciences, University of Novi Sad, Serbia 4Oncology Institute of Vojvodina, Sremska Kamenica, Serbia 5Centre for Dental Technology and the National Centre for Product Design and Development Research, Cardiff Metropolitan University, Cardiff, United Kingdom budaki@uns.ac.rs
Prejem rokopisa - received: 2011-07-14; sprejem za objavo - accepted for publication: 2011-08-24
The implementation of computer-aided technologies and systems has paved the way towards a significant advancement of the conventional modelling and manufacture of dental replacements. In this research the focus is on approach that combines reverse engineering as a modelling technique and rapid manufacture, i.e., selective laser melting, as the manufacturing technology, with special emphases on material selection in the fabrication of removable partial dentures. The paper presents the results of a biocompatibility analysis of the F75 Co-Cr dental alloy using the MTT eluate test.
Keywords: reverse engineering, rapid manufacturing, selective laser melting, removable partial dentures, biocompatibility, Co-Cr alloys, MTT eluate test
Uporaba računalniških tehnologij in sistemov je odprla pot do pomembnega napredka konvencionalnega modeliranja in izdelave snemljivih zobnih nadomestkov. Težišče te raziskave je približek, ki kombinira obratno inženirstvo kot tehniko modeliranja in hitre izdelave z laserskim taljenjem kot izdelavno tehnologijo s posebnim poudarkom na izbiri materiala za snemljive dele zobovja. V članku so predstavljeni rezultati analize biokompatibilnosti zobne zlitine F75 Co-Cr z MTT-preizkusom eluiranja.
Ključne besede: obratno inženirstvo, hitra izdelava, selektivno lasersko taljenje, snemljivi deli zobovja, biokompatibilnost, Co-Cr zlitina, MTT-preizkus eluiranja
1 INTRODUCTION	systems, CAD and reverse engineering (RE), CAE,
CAM, rapid manufacture (RM) (or additive manufacture
Dental prosthetics (also known as prosthodontics) has	that become the adopted term in the sector) and rapid
always maintained close relationships with engineering	prototyping (RP). The development and implementation
disciplines, relying mostly on production engineering.	of such technologies and systems have paved the way
The rapid development of computer-aided (CA) techno-	towards a significant advancement in conventional
logies, which completely transformed production engi-	modelling, manufacture and the inspection of dental
neering, also left an indelible mark on dental prosthetics.	replacements.1-6
Striving towards its primary goal - primum non nocere Different dental substrates may have special
(in English, 'Above all, do not harm!'), the area of dental	requirements related to their modelling and manufacture.
prosthetics has introduced numerous novel technologies	Removable partial dentures (RPDs) represent a special
and methods that allow the manufacture of precision,	type of denture, designed for partially edentulous
custom-made, optimal dental replacements. During the	patients who cannot have a fixed partial denture, i.e., a
past decade, efforts have been concentrated towards an	bridge. This type of prosthesis is referred to as
advancement of the modelling and manufacture of dental	removable, as patients can remove and reinsert it when
replacements by introducing modern CA systems,	required without professional help. Traditionally, RPD
state-of-the-art materials and machining technologies, as	frameworks are manufactured through the so-called
opposed to the traditional way of manual manufacture,	lost-wax technique, where a wax pattern burns out in a
which is prone to numerous subjective errors.1 Amongst	preheating unit followed by an immediate casting of the
modern CA systems that have found broad application in	melted alloy. Though in use for decades, this technique is
this area, the most widely used are 3D-digitization	sensitive and prone to human-induced errors.1,6
In this research the focus is on an approach that combines RE as a modelling technique, and RM i.e., selective laser melting (SLM), as the manufacturing technology, with a special emphasis on material selection in the fabrication of RPD frameworks.
The virtual design of dental restorations today almost always requires the application of RE modelling. RE, a modelling technique widely used in different engineering fields, has been increasingly applied in the field of prosthodontics during the past several years, mainly because of the rapid development of dental 3D digitization systems and the corresponding modelling software2,3. RM is no exception and its results in the field of prosthodontics significantly depend on RE modelling. Realizing the benefits of RE and RM, recently there have been several research works related to the possible use of these technologies in the design and manufacturing of RPD frameworks.1'3'7-9
SLM, an RM technique, is based on a layer-wise material addition that allows the generation of complex 3D parts by selectively melting successive layers of metal alloy powder on top of each other. As presented by Eyers and Dotchev in5 and Vandenbroucke and Kruth in,10 SLM is very suitable for dental and medical applications, due to the complex geometry of the produced parts. A pilot study presented by Williams et al.9 showed that an RPD framework produced by SLM techniques was comparable to conventional frameworks in terms of accuracy, quality of fit and function. However, this conclusion is based on a single study and much work needs to be completed before a final conclusion can be reached.
A very important issue in prosthodontics is the material used, i.e., the alloy.10-13 This issue is even more important in RPD framework manufacturing as the application is limited to cobalt-chromium (Co-Cr) alloys due to the low rigidity of titanium (Ti) alloys, as described by Aridome et al. in14, and the unfavourable characteristics of gold (Au) alloys, that are otherwise widely used in the fabrication of other dental restorations. The application of SLM in the manufacturing of RPD, brings some additional material requirements related to mechanical properties1014 and biocompatibility.1516 Although a dental prosthesis fabricated by SLM showed the potential of SLM as manufacturing technique, there are still very few reports of SLM application in the manufacture of RPDs from Co-Cr alloys. Though Co-Cr alloys have been used in dentistry for years, little is known about the influence of the SLM process on the alloys' biocompatibility and mechanical behaviour.
This paper, with regards to the above discussion, focuses on the applicability and possible benefits of the application of RE and RM techniques in the design and manufacture of RPD frameworks. Moreover, special attention in this research has focused on the material properties related to SLM application in the manufacture
of RPD frameworks. The paper also presents the results of an analysis of biocompatibility conducted with an MTT eluate test of the F75 Co-Cr dental alloy.
2 RE/RM IN DESIGN AND MANUFACTURING OF RPD FRAMEWORKS
The RE and RM combination has been recognized as fully compatible and very effective. Potential advantages include: a decrease of manufacturing time, an inherent repeatability, and the achievement of high quality through eliminating operator variations that are usually connected with the conventional (manual) design and manufacture of an RPD (Figure 1).
The application of RM, i.e., SLM in an RPD framework fabrication implicates the workflow presented in Figure 2. This workflow clearly presents three main phases:
1.	RE modelling (of the patient's cast),
2.	virtual design of the RPD framework,
3.	RM of RPD framework.
The first two phases are frequently unified and denominated in references as the CAD phase, while the third is often described as the CAM phase.1'3'5'7-10
The RE phase starts with the 3D digitization of the patient's cast. This usually includes acquiring a dental
Figure 1: Conventional (manual) design of RPD Slika 1: Konvencionalno (rocno) načrtovanje RPD
Figure 2: The typical workflow of an RE/RM approach in RPD framework fabrication
Slika 2: Značilen tok RE/RM približka v okviru RPD-izdelave
impression and extra oral scanning of a gypsum model produced from the impression (Figure l).1,6 However, the process could replace the need for an impression by the application of intra-oral scanning2 or CT.6,17,18 The point cloud obtained almost always needs to be pre-processed in order to insure a high-quality surface reconstruction, i.e., a credible CAD model. Regarding the applied 3D digitization technique/system, the pre-processing step can include different processes, such as noise filtering, data reduction, segmentation of the point cloud parts or assembling.19 The obtained surface model (the reference model or the "buck") is usually exported to an STL file format based on triangular
polygons, which is a suitable format for virtual dental surveying and virtual sculpting environments.8
The initial step of the RPD framework-design phase is the virtual dental surveying that is needed to identify areas of undercut present on the CAD model of the patient's teeth and soft tissues (Figure 3a). Unwanted undercuts have to be removed in order to ensure an unobstructed withdrawal of the RPD from the patient's orifice.1,8 However, there are some useful undercuts that need to be retained and their identification and measurement are important as they serve as secure holders of flexible clasps that provide reliable retention.1 The next step is the modelling of reliefs - the parts of a model that prevent the RPD framework from resting on the surfaces of soft tissues (Figure 3b).1 After the reliefs have been added, virtual sculpting of the RPD framework elements (1-occlusal rest, 2-polymeric retention frame, 3-lingual bar, 4-acrylic line, 5-non-active clasp, 6-guide plate) may begin (Figure 3c).
The virtual sculpting stage is based on software tools enabling analogous work to that used in physical sculpting. This is enabled through a haptic interface that incorporates positioning in 3D space and allows rotation and translation in all axes, transferring hand movements into the virtual environment (Figure 4). Moreover, haptic systems enable the operator to feel contact with the object that is the subject of the modelling. Besides this usage of the haptic arm in a freehand manner, the process of virtual sculpting also allows the application of standard CAD parametric features based on sizes, shapes, relations and positions.1,8
Once the CAD model of the RPD framework is obtained, it can be passed to the SLM after its preparation in appropriate software. This preparation primarily involves the creation of an adequate support (Figure 5) that acts as a firm base for the RPD framework to be built onto and which also conducts heat away during the sintering processes. As the supports need to be removed after the solidification of the part, it is important to avoid placing them on the fitting surface of the RPD.8
Figure 3: RPD framework design phase - the main steps: a) Identification of undercuts, b) Relief modelling, c) RPD framework elements Slika 3: Okvir faze RPD-načrtovanja - glavni koraki: a) identifikacija spodnjih prerezov, b) modeliranje reliefa, c) elementi okvirja RPD
Figure 4: Virtual design of RPD frameworks Slika 4: Virtualno na~rtovanje podlag RPD
Figure 5: Support in SLM manufacturing of RPD frameworks Slika 5: Podlaga pri SLM-izdelavi RPD-podlage
Figure 6: Principle of the SLM process Slika 6: Princip SLM-procesa
During the SLM process, powdered material is spread by a hopper and wiper mechanism. To accommodate a new layer of the material, the build platform has to move down by one layer thickness. Subsequently, the powder is deposited incrementally on top of each solidified layer, and the process is repeated (Figure 6).
The manufactured RPD framework needs to be finished and polished and this is performed using traditional dental laboratory procedures.8,9 Finally, the finished RPD framework has to be evaluated on the patient's cast. This is performed by a prosthodontic
expert, who will assess the quality of fit according to recommended practice.8
3 BIOCOMPATIBILITY TESTING OF THE SLM Co-Cr ALLOY F75 (BY MTT ELUATE TEST)
One of very important issues that need to be investigated is the biocompatibility of the specific Co-Cr alloy used for SLM, since - to the best of authors' knowledge - there are no known published conclusions about biocompatibility. Though the basic chemical elements in alloys used for SLM (F75) and conventional investment casting, also known as the lost-wax technique (Remanium 380+) generally match, they differ by a small percentage due to the specific requirements needed by the process (Table 1). However, it has been shown that a modification in composition and pre-treatment can influence the cytotoxicity of an alloy on a large scale.20,21
The previous discussion motivated the authors to start research related to the biocompatibility testing of the specific Co-Cr alloy used for SLM.
In biocompatibility evaluations of alloys, cell culture studies are the usual starting point as they enable an investigation of the toxicity in a simplified system that minimizes the effect of confounding variables.21 Thus, within this study a murine fibroblast cell line (L929) was used in accordance with the requirements of the ISO standard 7405 (ISO 2008).15 The cytotoxicity determination of the Co-Cr alloy used for the fabrication of an SLM RPD framework was based on the MTT eluate test method.
3.1 Sample preparation
The investigation included the fabrication of two groups of test samples - the first from the SLM and the second from conventional investment casting (Figure 7).
Figure 7: Test samples obtained by SLM (left) and by vacuum casting (right)
Slika 7: Preizkusanci SLM (levo) in vakuumskega taljenja (desno)
Table 1: Composition of the Remanium GM 380+ and Sandvik Osprey F75 alloys in mass fractions, w/% Tabela 1: Sestave zlitin Remanium GM 380+ in Sandvik Osprey F75 v masnih deležih, w/%
Ingredients, w/%	Co	Cr	Mo	Si	Mn	N	C	Fe	Ni
Remanium GM 380+	64.6	29	4.5	<1	<1	<1	<1	/	/
Sandvik Osprey F75	Balance	27-30	5-7	<1	<1	/	<0.35	<0.75	<0.5
Figure 8: SLM system - Realiser MTT-Group, UK Slika 8: SLM-sistem Realiser MTT-grupe, VB
The first group of samples was manufactured by the SLM system Realiser MTT-Group (Figure 8) and the software was Magics 9.5, Materialise NV. The Co-Cr layer thickness was 0.075 mm, the laser's maximum scan speed was 300 mm/s, and the beam diameter was 0.150-0.200 mm. The F75 Co-Cr alloy (Sandvik Osprey Ltd., UK) used in this study is a spherical powder with a maximum particle size of 0.045 mm (particle size range 0.005-0.045, mean size approx. 0.030 mm). After the SLM process was completed and specimens' supports were removed, the discs were finished by polishing according to the usual dental laboratory procedure.
The second group of samples - four disc specimens with a radius of 5 mm and thickness of 1 mm - were fabricated from a non-precious Co-Cr alloy Remanium GM380+ (Dentaurum, Ispringen, Germany) containing no Ni, Be or Fe, and widely used for RPD framework casting. The discs were obtained from wax patterns,
invested in Rema dynamic (Dentaurum, Ispringen, Germany) investment, and vacuum casted in a Nautilus CC (Bego, Bremen, Germany) system (Figure 9). After casting, the discs were divested and blasted with 100-pm aluminium oxide particles, and then polished with silicon carbide papers in the sequence 320, 400, 600, 1200, 1500 and 2000. The final polishing was performed using oxide pastes.
3.2 MTT eluate testing
The test performed was the MTT (tetrazolium colori-metric assay) eluate test, widely used for the quantitative evaluation of cell proliferation and survival.16 20 The assay depends on the cleavage of tetrazolium salt 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazoliumbr omide (MTT) to purple formazan crystals by mito-chondrial dehydrogenases in viable cells.21 The assay detects living but not dead cells and the rate of MTT reduction to formazan products. It is also dependent on the degree of cell activation. Therefore, the assay is suitable for measuring cytotoxicity, proliferation and activation.16
Eluates of both the CM and SLM disc samples were prepared. The samples were extracted with 10 mL of Dulbecco's modified Eagle's medium (DMEM) without serum for 48 h. The extraction was performed in an atmosphere of 5 % CO2 and 95 % air at 37 °C. All the extracts were filtered for sterilization and used as a culture medium for L-929 cells. MTT assays with L-929 cells treated with different eluates were completed after 48 h of incubation. The experiment was repeated twice (thus eight CM and eight SLM samples were tested in two independent experiments).
The cells (L929) were cultured in Petri dishes containing eluates of the CM or SLM alloy discs (Figure 10). They were incubated for (3, 5, 7 and 9) d at 37 °C in 95 % air and 5 % CO2. The control samples contained a regular culture medium. After the incubation, the cells were detached using enzymatic digestion and counted in a counting chamber using trypan blue. Briefly, 5 x 103 cells were seeded to a 96-well plate and cultured for 48 h at 37 °C in 95 % air and 5 % CO2.
After the incubation period, 20 pL of MTT solution were added to each well and incubated for another 3 h. The purple formazan product was dissolved in 100 pL of
Figure 9: Nautilus CC system for vacuum casting
Slika 9: Nautilusov sistem za vakuumsko taljenje
Figure 10: MTT eluate testing Slika 10: MTT preizkus eluiranja
Figure 11: MTT test results for CM and SLM alloy after different incubation periods - graphical review of the confidence intervals Slika 11: Rezultati MTT-preizkusov CM- in SLM-zlitine po različnih dobah inkubacije - grafična predstavitev intervalov zanesljivosti
0.04-M hydrochloric acid in isopropanol. The reduced MTT was then measured spectrophotometrically in a dual-beam, microtiter plate reader Multiscan MCC/340 at 540 nm with a 690-nm reference. The optical density values of the experimental groups were divided by the control and expressed as a percentage of the control.
3.3 Statistical results of the MTT eluate testing
A statistical analysis was carried out using the Statgraphics Centurion program. The data were evaluated statistically using the Student's t-test and a value of p < 0.05 was considered to be statistically significant. The results of the MTT eluate testing of the disc samples that were taken after an extraction period of (3, 5, 7 and 9) d are listed in Table 2, and graphically presented in Figure 11.
Of particular interest was the confidence interval evaluation for the difference between the means. Since all four intervals contain the value 0, there is no statistically significant difference between the means of the CM and SLM samples at the 95 % confidence level. Furthermore, a t-test was used for testing a specific hypothesis about the difference between the means of the populations from which the two samples come. In this case, the test was constructed to determine whether the difference between the two means equals 0.0 (null hypothesis: meanl = mean2) versus the alternative hypothesis that the difference does not equal 0.0 (meanl ^ mean2). Since the computed P-values are not less than 0.05 in all four cases, the null hypothesis cannot be rejected.
4 DISCUSSION
Publications investigating the corrosion of dental alloys give information firstly about the release of potentially harmful ions from the dental device. Cell-culture tests give an insight into whether the released ions in a cell-culture medium imitating the oral environment could have a negative effect on the biological system. The results of the MTT eluate testing with the SLM samples do not show significant cellular damage potential. Statistical analyses carried out showed that the alloys did not release harmful material that could cause acute effects against L929 cells under the given experimental conditions. Furthermore, the MTT test showed no permanent damage to the cell function. The viability was much higher than 50 % after all the extraction periods for both the CM and SLM alloy. Replication during an extended contact period with
Table 2: Results of statistical analysis of MTT eluate testing of CM and SLM disc samples Tabela 2: Rezultati statistične analize MTT-preizkusa eluiranja CM- in SLM-vzorcev
	Period of cell incubation in d	3		5		7		9	
	Technology	CM	SLM	CM	SLM	CM	SLM	CM	SLM
CO CJ	Count	8	8	8	8	8	8	8	8
c^ .£3	Average % of relative cell No.	104.655	103.639	107.604	108.364	113.73	112.694	119.778	120.525
CD o CO CD	Standard deviation	3.62557	4.56325	3.43234	2.94701	3.34183	2.92113	3.45744	3.58149
	Coeff. of variation,%	3.4643	4.40303	3.18979	2.71955	2.93839	2.5921	2.88655	2.97158
Q	Minimum	100.05	98.64	102.04	101.54	109.78	109.78	114.01	115.61
	Maximum	109.45	111.23	111.97	111.02	119.74	117.64	124.31	125.64
O C o CO	95 % CIM*	104.655 +/-3.03106	103.639 +/-3.81498	107.604 +/-2.86951	108.364 +/-2.46377	113.73 +/-2.79385	112.694 +/-2.44213	119.778 +/-2.8905	120.525 +/-2.99421
(S O	95 % CIDM**	1.01625 +/-4.41952		-0.76 +/-3.43048		1.03625 +/-3.36576		-0.7475 +/-3.77485	
U	t	0.493186		-0.475165		0.660339		-0.424715	
	P-value	0.629527		0.641997		0.519754		0.6775	
* CIM - Confidence Interval for Means, ** CIDM - Confidence Interval for the difference between the Means assuming equal variances
potential toxic substances, however, showed good biocompatible properties of the chosen SLM alloy. Additionally, the negative effect decreased with time for both the examined substances. Therefore, both alloys can be rated as non-cytotoxic.
It has, however, to be noted that SLM, as a complex thermo-physical process, produces a variation in the final product depending on several factors, such as the material, laser, scan and parameters of the environment used.10 Changeable variables include: laser power, layer thickness, scan speed and hatch spacing. With the current settings, as can be seen from the study, the final product complies with the required biocompatibility standards, showing no potentially harmful effect. However, those values can be adjusted accordingly, optimizing some aspects that can have a negative effect on the materials' properties, such as porosity. For example, for a low energy input, successive scan tracks may not be fully molten, leaving large pores along the scan lines, as seen in the mentioned study. If so, a combination other parameters might change the surface properties, which might also result in changes in the ion release and therefore require separate biocompatibility studies. This study has, however, showed that the initial screening gave positive results and the F75 SLM alloy can be subjected to further tests.
The study concerning the ion release from the cast and SLM samples, presented in,10 revealed the more favourable behaviour of the SLM specimens. The main ion detected was cobalt, since the corrosion of the alloy is determined by the main component, and the passivating effect of chromium. The SLM test specimens showed lower emissions than the cast specimens, probably because the laser-melted material is more homogeneous, contains fewer pores and has a finer microstructure. This also highlights the importance of the finishing procedure, which still has to be conducted manually.
Another physical factor that might influence the biocompatibility 'in vivo' is the surface roughness21. The changes in this parameter can be explained by the so-called stair effect, inherent to the layer-wise production of SLM. In the oral environment, this might increase plaque retention, leading to the formation of acidic micro-fields that might change the metallic ion release unfavourably. This effect can be reduced by decreasing the layer thickness or by increasing the sloping angle10.
5 CONCLUSION
This paper shows that a complete RE/RM procedure for RPD framework fabrication should bring significant
advantages both to practitioners and patients. Special attention was focused on the biocompatibility analysis of the dental alloys used with the SLM. Moreover, the paper presents a biocompatibility evaluation of the F75 SLM dental alloy using the MTT eluate test. On the basis of the obtained results, within the limitations of the study, it can be concluded that the RE/RM procedure showed a promising potential in RPD framework fabrication, as well as that the F75 alloy used for SLM manufacturing showed positive initial results regarding its biocompatibility. However, further studies, including in vivo tests and tests of mechanical properties, have to be conducted before the final release of the alloy for mass production.
6 REFERENCES
1	D. Eggbeer, R. Bibb, R. Williams, Proc. Inst. Mech. Eng. Part H-J. Eng. Med., 219 (2005) 3, 195-202
2	I. Budak, B. Trifkovic, T. Puskar, M. Hadzistevic, D. Vukelic, J. Hodolic, Application and accuracy of 3D-digitization systems in the field of dentistry, Proceedings of the 6th International Working Conference Total Quality Management - Advanced and Intelligent Approaches, 2011, 409-414
3M. Germani, R. Raffaeli, A. Mazzoli, Rapid Prototyping J., 16
(2010)	5, 345-355
4 A. Cernescu, N. Faur, C. Bortun, M. Hluscu, Eng. Fail. Anal., 18
(2011)	5, 1253-1261
5D. Eyers, K. Dotchev, Assem. Autom., 30 (2010) 1, 39-46
6	A. Azari, S. Nikzad, Rapid Prototyping J., 15 (2009) 3, 216-225
7	R. Bibb, D. Eggbeer, R. J. Williams, Rapid Prototyping J., 12 (2006) 2, 95-99
8R. J. Williams, R. Bibb, D. Eggbeer, J. Collis, J. Prosthet. Dent., 96 (2006) 2, 96-99
9R. J. Williams, R. Bibb, D. Eggbeer, Pract. Proced. Aesthet. Dent., 20 (2008) 6, 349-351
10	B. Vandenbroucke, J. P. Kruth, Rapid Prototyping J., 13 (2007) 4, 196-203
11	R. Rudolf, T. Zupancic Hartner, L. Kosec, A. Todorovic, B. Kosec, I. Anzel, Metalurgija, 47 (2008) 4, 317-323
12	K. Raic, R. Rudolf, A. Todorovic, D. Stamenkovic, I. Anzel, Mater. Tehnol., 44 (2010) 2, 59-66
13A. Todorovic, K. Radovic, A. Grbovic, R. Rudolf, I. Maksimovic, D.
Stamenkovic, Mater. Tehnol., 44 (2010) 1, 41-47 14 K. Aridome, M. Yamazaki, K. Baba, T. Ohyama, J. Prosthet. Dent.,
93 (2005) 3, 267-273 15ISO 7405:2008, Dentistry - Evaluation of biocompatibility of medical devices used in dentistry.
16	ISO 10993-5:2009, Biological evaluation of medical devices - Part 5: Tests for in vitro cytotoxicity.
17	R. Bibb, D. Eggbeer, P. Evans, A. Bocca, A. Sugar, Rapid Prototyping J., 15 (2009) 5, 346-354
18	R. Bibb, J. Winder, Radiography, 16 (2010) 1, 78-83
19	I. Budak, M. Sokovic, J. Kopac, J. Hodolic, Strojniski Vestn. - J. Mech. Eng., 55 (2009) 12, 755-765
20	T. Mosmann, J. Immunol. Methods, 65 (1983) 1-2, 55-63
21	A. Naji, M. F. Harmand, J. Biomed. Mater. Res., 24 (1990) 7, 861-871