Jacobs Journal of Medical Diagnosis and Medical

OPEN ACCESS

Jacobs Journal of Medical Diagnosis and Medical Imaging Research Article

A Tissue-Equivalent Phantom for the X-ray Evaluation of Polymeric Orthopedic Implants with Osseointegrative Coating Francesco Tessarolo1,2*, Nicola Pace2, Gloria Miori3,4, Giandomenico Nollo1,2, Mauro Recla5, Aldo Valentini3

Department of Industrial Engineering, University of Trento, Trento, Italy

1

Healthcare Research and Innovation Program (IRCS-FBK-PAT), Bruno Kessler Foundation, Trento, Italy

2

Department of Health Physics, Azienda Provinciale per i Servizi Sanitari, Trento, Italy

3 4

School of Medical Physics, University of Roma Tor Vergata, Rome, Italy

Division of Diagnostic Radiology, S. Chiara Hospital, Trento, Italy

5

*Corresponding author: Dr. Francesco Tessarolo, Healthcare Research and Innovation Program (IRCS-FBK-PAT), Bruno Kessler Foundation & Department of Industrial Engineering, University Of Trento via delle Regole, 101, I-38123, Mattarello, Trento, Italy, Phone: +39 0461 282775; Fax: +39 0461283659; E-mail: [email protected], [email protected] Received: 09-13-2016 Accepted: 10-18-2016 Published: 11-03-2016 Copyright: © 2016 Francesco Tessarolo Abstract Purpose Orthopaedics polymeric implants made of ultra-high molecular-weight polyethylene (UHMWPE), polyether-ether-ketone (PEEK) or its carbon-fiber reinforced composite (CFR-PEEK) can be coated with titanium (Ti) or hydroxyapatite (HA) by compression-moulding or plasma-spray techniques for bone integration. The coating may also improve the radiological properties of the implanted device. This work aimed at developing a tissue-equivalent phantom for measuring the radiological contrast given by different polymeric materials coated with Ti or HA for orthopedic implant applications.

Methods A modular phantom was designed to simulate the human hip, differentiating soft tissue, trabecular and cortical bone. Seven different substrate-coating combinations were considered for X-ray contrast measurement. Substrate thickness was also varied resulting in 28 tested coupons. Radiographic images were obtained by a state-of-the-art clinical radiography unit. Contrast values were compared to a threshold for detectability. Results The tissue equivalent phantom allowed rapid coupon exchange and acquisition of multiple X-rays-images in a short period of time. Coupons contrast ranged from 0.02 to 0.27 depending on the combination of polymer type, substrate thickness, coating composition, thickness and porosity. Eighteen coupons showed a contrast higher than the predefined threshold. Thin titanium coatings resulted in a sufficient contrast when deposited on PEEK or CFR-PEEK thicker than 10 mm. Medium thickness coatings with low porosity and thick coatings with high porosity were above the threshold when associated to polymeric substrates with a thickness of 8 mm or more. Differently, pre-sintered grid made of titanium alloy was clearly distinguished even when associated to only 6 mm thin UHMWPE. Cite this article: Francesco Tessarolo. A Tissue-Equivalent Phantom for the X-ray Evaluation of Polymeric Orthopedic Implants with Osseointegrative Coating. J J Med Diagnosis and Image. 2016, 2(1): 014.

Jacobs Publishers

Conclusions The proposed phantom is a valid tool to assess coated polymeric implants by X-ray plain radiography. Considering the wide range of coatings and polymers nowadays available, this tissue-equivalent phantom could help in optimizing micro and macro design of implant components. Keywords: X-rays imaging; Prosthetic implant; Contrast; Tissue-equivalent phantom; Hip

Abbreviations

UHMWPE: Ultra-High Molecular Weight PolyEthylene; PEEK: PolyEther-Ether-ketone;

CFR/PEEK: Carbon-Fiber-Reinforced PolyEther-Ether-Ketone; Ti: Titanium;

HA: HydroxyApatite;

AP: Antero-Posterior;

CT: Computed Tomography;

NIH: National Institutes of Health; S: Sample; B: Blank;

Ti-6Al-4V: Grade 5 Titanium Alloy;

ASTM: American Society for Testing Materials; DR: Digital Radiography;

SDD: Source to Detector Distance; ROI: Region of Interest; G: Grey Value;

PI: Pixel Intensity; ∆PI: contrast

Introduction Since the pioneristic prosthetic hip joint implanted by Sir John Charnley, the role of polymeric components in artroprosthetic replacement has been fundamental [1]. Ultra-high molecular weight polyethylene (UHMWPE) contributed significantly to the success and diffusion of total joint replacements in the last 50 years. This polymer has good tribological properties, but it is unable to integrate with bone tissues. Metallic shells or acrylic cement is needed to guarantee mechanical coupling be-

2 tween UHMWPE implant and bone [2]. Unfortunately, friction at the interface may result in wear debris and chronic reaction in periprostetic tissues, eventually leading to osteolysis and implant loosening [3-6]. In case of metal-backed cups, the friction between acetabular liner and metal shell may also be a source of polyethylene debris with negative consequences on long-term stability of the prosthesis [7,8]. In order to minimize the backside wear, new monoblock cups were recently developed, consisting of a UHMWPE liner directly coupled to a metal shell or coated with titanium particles by means of compression moulding techniques [9]. Showing higher mechanical and tribological properties in respect to UHMWPE [10-12], Polyether-ether-ketone (PEEK) and its carbon-fiber-reinforced composite (CFR/PEEK) have been increasingly employed as biomaterials for trauma, orthopedic, and spinal implants. PEEK and its composites can be plasma-sprayed with powders of titanium (Ti), titanium alloys or hydroxyapatite (HA) to obtain primary implant stability and osseointegration [13]. Thermal characteristics and mechanical properties of PEEK polymers, added to their chemical stability in the biological environment, gave rise to the possibility of designing new implants for arthroprosthetic applications (e.g. acetabular cups, finger joint implant) and arthrodhesis (e.g. polymeric cages for spinal fusion) [12].

For all these prosthetic devices in which a bulky polymeric component is directly coupled or coated with a metallic or ceramic layer, the surface treatment have also an impact on the radiographic properties of the component. From the clinical standpoint, a relevant issue in orthopedic surgery is the possibility to image the implant both intra-operatively and at follow up, in order to assess the component migration or loosening in the early and late postoperative periods [14]. To perform such evaluation, plain radiography is widely used, providing the physician with qualitative and quantitative information about the location and orientation of the implant in a non-invasive way [15-17]. Polymers, as medical grade UHMWPE and PEEK have scarce opacity to X-ray investigation, resulting in X-rays transparency. Although PEEK radiolucency offers some advantages for the radiographical evaluation of interbody spinal fusion devices [18], many other orthopaedics applications, including hip and knee arthroplasty, take advantage from radiopaque components [16]. The presence of a osseointegrative coating to a polymeric implant could enhance the X-rays contrast of the device. Since coating thickness, porosity, and roughness can be finely tuned by properly adjusting coating parameters, powder type and grain size [13], according to clinical application and patient's needs [12, 19], it is of utmost importance to develop a quantitative methodology to characterize the X-rays contrast of the coated component in order to predict the radiological behavior of the implant, once it is inserted into the patient's body. Considering that, in literature, methodologies for the radiographic characterization of coated polymeric components for hip prostheses are lacking, this work aimed at developing

Cite this article: Francesco Tessarolo. A Tissue-Equivalent Phantom for the X-ray Evaluation of Polymeric Orthopedic Implants with Osseointegrative Coating. J J Med Diagnosis and Image. 2016, 2(1): 014.

Jacobs Publishers and testing a tissue-equivalent phantom for quantifying the radiological contrast of coated polymeric components.

3

Materials and Methods

Phantom Design and Construction A X-rays tissue-equivalent modular phantom was designed to mimic a location for hip arthroprosthetic implant. The antero-posterior (AP) projection of the pelvis was selected as the most frequently applied X-rays protocol for planning and checking the position of the acetabular cup implant [14,15]. Tissue-equivalent plastic materials (Computerized Imaging reference Systems, Inc. Norfolk, VA, USA) reproducing X-rays absorption and scattering properties of three representative human tissues of the hip joint (soft tissue, trabecular bone and cortical bone) in the 10 keV-100 MeV energy range were used to realize the phantom. The anatomical thickness of soft tissue, trabecular bone and cortical bone across the hip joint along the sagittal body plane was obtained by pelvis computed tomography (CT) images available from the freely accessible Human Visible Project Database (NIH) (http://www.nlm.nih. gov). Phantom dimensions are summarized in (Table 1). Tissue type

Thickness (mm)

Soft tissue

110.0 ± 0.5

Trabecular bone tissue

25.0 ± 0.2

Cortical bone tissue

10.0 ± 0.2

Table 1. Thicknesses of tissue-equivalent materials in the hip phantom.

The phantom consisted of four superimposed soft tissue equivalent slabs (40 mm in thickness). The two medial slabs presented two parallel cylindrical cavities (40 mm in diameter, 80 mm in height), able to host a modular combination of insets realized with trabecular and cortical bone equivalent tissues (Figure 1). Two symmetric sample holders were realized in each trabecular bone insert to allow placement and rapid exchange of the samples under investigation (disk-shaped coupons, 25.4 mm in diameter and up to 20 mm in thickness). The first sample holder (S) was used to lodge the coupon under testing and the second holder (B) was left unloaded as blank. Prosthetic Implant Coupons

Twenty-eight disk-shaped coupons (25.4 mm in diameter) representative of different polymeric coated prosthetic components were studied for their radiological characteristics by means of the tissue-equivalent phantom.

Figure 1. Design of the phantom indicating the different type of tissue equivalent materials used. (a) latero-lateral (top) and antero-posterior (bottom) projections. S and B indicate respectively the cavities for the sample placement and blank measurement. Measures are in cm.


Jacobs Publishers The coupons substrates were made of three commercially available medical-grade polymers: GUR® 1050 (UHMWPE) supplied by Quadrant EPP (Milan, Italy), PEEK-Optima® LT1 (unfilled PEEK), and Motis® (CFR-PEEK) supplied by Invibio Ltd. (Lancashire, UK). The three substrates were combined with different coatings (hydroxyapatite, titanium, titanium alloy) to obtain seven coating-polymer combinations of interest for orthopedic implant applications. The final number of coupons was obtained by further varying the thickness of the polymeric substrate in four different values (6, 8, 10, and 12mm) representing the common range of thickness for polymeric implant components in the hip.Coupons were prepared and coated by the company Eurocoating S.p.A. (Trento, Italy). PEEK and CFR-PEEK polymers were coated by plasma spray techniques while coatings on UHMWPE were obtained by thermal compression of titanium particles or of a pre-sintered three-dimensional Ti-6Al-4V grid on the polymer surface. The coatings presented different thickness and porosity according to different process conditions and powder characteristics. Samples properties are summarized in Table 2. Coupon name

Substrate

4

Image Acquisition and Analysis

Contrast measurements were performed by placing each single coupon in the sample holder (S) of the phantom and leaving the blank holder (B) empty.

Radiographic contrast of samples in the tissue equivalent phantom were evaluated by applying a standard clinical im age acquisition protocol for the antero-posterior pelvis (Tube voltage: 81 kVp, SDD: 115 cm, with anti-scattering grid between samples and detector) on a clinical digital radiography (DR) unit (Axiom Aristos FX Plus, Siemens). The phantom was placed centrally in the field of view of the radiographic system avoiding the dose detection system (Automatic Exposure Control – ionization chambers) to prevent having different values of contrast detection throughout the phantom (Figure 2). Evaluation of Contrast Values ∆PI

Acquired images were converted into 8-bit format in order to Coating

Composition Thickness (mm)

Composition

Thickness (μm)

Porosity (%)

Mean ± SD

Mean ± SD

PEEK HA_69

PEEK

6, 8, 10, 12 ± 0.2

Hydroxyapatite

69± 7

8± 2*

PEEK Ti_47

PEEK

6, 8, 10, 12 ± 0.2

Titanium

47± 13

8± 1*

CFR-PEEK Ti_52

CFR-PEEK

6, 8, 10, 12 ± 0.2

Titanium

52± 13

26 ± 1*

CFR-PEEK Ti_145

CFR-PEEK

6, 8, 10, 12 ± 0.2

Titanium

145± 47

34 ± 4*

CFR-PEEK Ti_317

CFR-PEEK

6, 8, 10, 12 ± 0.2

Titanium

317± 75

61 ± 4*

UHMWPE Ti_350

UHMWPE

6, 8, 10, 12 ± 0.2

Titanium

350± 120

39± 3**

UHMWPE TiAlV_2100

UHMWPE

6, 8, 10, 12 ± 0.2

Ti-6Al-4V alloy

2100± 1200

34± 3**

Table 2. Coating and substrate characteristics of coupons tested with the phantom. * Porosity was evaluated as % of air between metallic powders ** Porosity was evaluated as % of polymer between metallic powder

Coatings thickness and porosity were measured in accordance with standard test method ASTM F1854 2009 by metallographic preparation and image analysis on equivalent samples.

set the grey-scale of each pixel at 0 for black and at 255 for white, by means of ImageJ (NIH). Two equivalent circular regions of interest (ROIs) of about 104 pixels each were


Jacobs Publishers

5

Figure 2. Tissue-equivalent phantom in the open configuration (left) and in closed configuration under the X-rays beam (right) of the Axiom Aristos FX Plus, Siemens. In the open configuration, a coupon made by Ti-6Al-4V coating on 12 mm UHMWPE is loaded in the sample placement. Metallic pins at phantom corners were used to fasten the phantom modules. An Al stepwedge ranging from 12mm to 1mm (1 mm step) was also positioned in the field of view during image acquisition.

identified within the area of the two sample holders (ROIS for the sample; ROIB for the blank) and the mean grey-scale values were measured (Figure 3).

From the mean grey value (G), pixel intensity (PI) of the area of interest can be calculated by the following formula [20,21]:

Sample contrast (∆PI) was therefore calculated by the equation: where GS is the mean grey value of the sample within the phantom and GB is the mean grey value of the sole phantom. For each sample, standard deviation of the grey-scale values in the selected ROI was computed and considered as an indicator of measurement uncertainty.

Figure 3. Acquired X-ray image of samples in the tissue-equivalent phantom with the identification of the two region of interest (black circles) used for contrast computation. S: sample, B: reference (blank).

Contrast values were compared to the suggested value for minimum detectability in a radiographic medical image as reported in ASTM standards (ASTM F640 2007). The sample was considered to have a sufficient contrast if the mean value minus one standard deviation was above the ASTM threshold ∆PIT=0.05. Below this value, contrast was considered unsatisfactory to properly distinguish the sample in respect to blank (unloaded phantom).


Jacobs Publishers

Results The tissue equivalent phantom showed to be convenient to use, allowing a rapid coupon exchange and the acquisition of multiple X-rays-images in a short period of time. Beam alignment and reproducible positioning in respect to X-rays source and detector were easily achieved across different acquisition within the same working session and in different imaging sessions.

6 coupons with Ti coating on CFR-PEEK showed that radiological contrast was directly proportional to coating thickness and inversely to coating porosity, for the same substrate. This gave equivalent ∆PI values for Ti_145 and Ti_317 since higher thickness (leading to higher ∆PI) was compensated by higher porosity (leading to lower ∆PI).

Contrast values (∆PI) of tested coupons ranged from 0.02 to about 0.27 (Figure 4).

Figure 4. Representative X-ray images of the phantom loaded with coupons with high (left) and low (right) contrast value. S: sample, B: blank (empty sample holder). An Al stepwedge ranging from 12mm to 2mm (1 mm step) was included in the field of view of each image. Images were acquired with Axiom Aristos FX Plus, Siemens according to the standard clinical protocol for an antero-posterior pelvis imaging (beam energy: 81 kVp) with a source to detector distance of 115 cm and the anti-scattering grid positioned between sample and detector. No image post-processing has been performed.

The 6 mm thick PEEK coupons coated with hydroxyapatite showed the lower contrast among the tested samples. Conversely, the highest contrast value was associated to the Ti6Al-4V grid on the 12 mm thick UHMWPE. Intermediate contrast values were obtained for titanium coatings on different polymeric substrates. ∆PI values obtained from all the 28 test coupons are plotted in Figure 5.

Both the polymeric substrate and the coating contributed to the measured contrast of the coupons, depending on the combination of polymer type, substrate thickness, coating material, coating thickness and porosity. Polymer thickness significantly affected the contrast of the coupons, resulting in different ∆PI values among coupons with different substrate thickness but equal coating. Moreover, thick coating with medium porosity (e.g. Ti_350) on UHMWPE showed less contrast than thinner and less porous coatings (e.g. Ti_145) on CFR-PEEK. Data from

Figure 5. X-ray contrast values (∆PI) of the twenty eight test coupons grouped according to coating-substrate combination. Polymer substrate thickness (6, 8, 10, 12 mm) is also distinguished within the same group. The horizontal dashed line indicates the threshold of detectability. (∆PIT=0.05) according to ASTM 650. Coupons indicated with “*” were considered to have a sufficient radiographic contrast.

Measurement uncertainty was primarily related to the variability of thickness and porosity within the coating and, secondarily, to the intrinsic noise of the radiological system. The variability of radiological properties within the same coating was evident in the thicker titanium coatings obtained with a coarse grain metal powder.

The comparison with the ASTM threshold (∆PIT=0.05) revealed that only a fraction of coupons tested in this study would be visible by plain radiography when positioned in the hip. A sufficient radiographic contrast (i.e. above the settled threshold) was found in 18 out of the 28 tested coupons, as indicated in Figure 5. The hydroxyapatite coating (HA_70) on PEEK showed a sufficient contrast only when deposited on a 12 mm thick PEEK substrate. Thin titanium coatings (e.g. Ti_47 and Ti_52) resulted in a sufficient contrast only when deposited on PEEK or CFR-PEEK thicker than 10 mm. Medium thickness coatings with little porosity and thick coatings with higher porosity (e.g. Ti_145 and Ti_317 and Ti_350) were above the threshold only when associated to polymeric substrates with a thickness of 8 mm or more. Differently, the Ti alloy grid (e.g. Ti-6Al-4V_2100) was clearly distinguished even when associated to 6 mm thin UHMWPE.


Jacobs Publishers

Discussion The phantom here presented allowed to reproduce different tissue layers imaged by the AP X-rays projection of the hip, and to simulate X-rays interactions occurring in plain radiography of a real hip implant. To model the implant location, three tissue-equivalent materials were used, simulating soft tissue, trabecular and cortical bone. Considering the imaging parameters typically used in AP plain radiography of the hip, distinction among various kinds of soft tissue surrounding the joint (i.e. ligaments, muscular tissue, adipose tissue) was not implemented into the phantom because of the limited added value. Moreover, the relative percentage of muscular and fat tissue in the pelvis is subjected to high variability according to patients characteristics (e.g. gender, age, and size). To properly simulate average radiological characteristics of soft tissues, a tissue equivalent plastic with intermediate characteristics between fatty and muscular tissue was chosen. The tissue-equivalent phantom showed to be effective for quantifying the radiographic contrast of UHMWPE, PEEK and CFR-PEEK coupons with different osteointegrative coating and substrate thickness ranging from 6 to 12 mm. The presence of two sample holders within the phantom was essential to reduce uncertainty of contrast measurements. The absorption of the sample into the phantom was always compared to the absorption of the sole phantom within the same radiographic image. This allowed compensating for any eventual system instability (e.g. fluctuation of the X ray beam) during the acquisition of multiple images. The comparison of the measured contrast with a minimum detection value (∆PI = 0.05), allowed to easily discriminate samples with sufficient clinical detectability when placed into the hip. Since the polymer type, substrate thickness, coating material, coating thickness and porosity are all factors influencing implant X-rays interaction, the a priori computation of the overall sample contrast could be rather complex. The use of the tissue equivalent phantom we proposed allows a rapid discrimination of X-rays detectable samples within the hip. If the device proved to be undetectable, the insertion of metallic markers should be considered.

Clinical protocols for image acquisitions do not represent universal healthcare standards, as different hospitals and professionals may use slightly different acquisition parameters. For such a reason, although the clinical protocol considered in the present study for image acquisition represents a local standard for AP pelvis image, the acquisition parameters may be different in other hospitals or radiological units. Therefore, the radiographic contrast of samples might turn out to be unsufficient using different acquisition parameters. Nevertheless, when the general guidelines are followed, the radiographic contrast of samples should not differ significantly.

7

The interpretation of the radiographic contrast value obtained by the tissue equivalent phantom should take into consideration that all the samples analysed presented a bi-dimensional geometry perpendicular to the X-ray beam. This working configuration guarantees the measurement of the minimum contrast due to the sample into the phantom. Differently, non-perpendicular configurations would bring to an increment of the sample contrast. Since the majority of polymeric implant components have complex designs, the bi-dimensional geometry represents a conservative estimate of the implant contrast.

Conclusion

In conclusion the proposed phantom can represent a valid tool to assess detectability of polymeric coated implants by X-ray plain radiography. The phantom modularity represented a peculiarity of the proposed system, allowing to modify thickness and combination of tissue-equivalent slabs in order to reproduce a wide range of different implant locations.

Considering the importance of X-rays imaging in orthopedics, and the wide range of coatings treatments, the information obtained from the tissue-equivalent phantom could be relevant in optimizing micro and macro design of polymeric coated implants.

Acknowledgments

The authors are grateful to Gianluca Zappini of Eurocoating S.p.A. (Trento, Italy) for having provided the coupons.

The study has been financially supported by Eurocoating S.p.A. through a grant of The Autonomous Province of Trento (project “Inspired”).

References

1. Charnley J. Surgery of the hip-joint: present and future developments. Br Med J. 1960, 1(5176): 821-826. 2. Kurtz SM. UHMWPE Biomaterials Handbook, 2nd ed. Elsevier, Inc; 2009.

3. Willert HG, Bertram H, Buchhorn GH. Osteolysis in alloarthroplasty of the hip. The role of ultra-high molecular weight polyethylene wear particles. Clin Orthop Relat Res. 1990, 258: 95-107. 4. Harris WH. The problem is osteolysis. Clin Orthop Relat Res. 1995, 311: 46–53. 5. Catelas I, Wimmer MA, Utzschneider S. Polyethylene and metal wear particles: characteristics and biological effects. Semin Immunopathol. 2011, 33(3): 257-271.


Jacobs Publishers 6. Krenn V, Hopf F, Thomas P, Thomsen M, Usbeck S et al. Supramacroparticulate polyethylene in inflammation of synovial-like interface membranes: Characterization and suggested nomenclature. Orthopade 2016, 45(3): 256-264. 7. Civinini R, D’Arienzo M, Innocenti M. A ten-year follow-up of the Reflection cementless acetabular component. J Bone Joint Surg (Br). 2008, 90(5): 570-573.

8. Krieg AH, Speth BM, Ochsner PE. Backside volumetric change in the polyethylene of uncemented acetabular components. J Bone Joint Surg Br. 2009, 91(8):1037-1043.

9. Weiss RJ, Hailer NP, Stark A, Kärrholm J. Survival of uncemented acetabular monoblock cups. Acta Orthop. 2012, 83(3): 214-219. 10. Wang A, Lin R, Polineni VK, Essner A, Stark C et al. Carbon fiber reinforced polyether ether ketone composite as a bearing surface for total hip replacement. Tribology International. 1998, 31(11): 661-667.

11. Wang A, Lin R, Stark C, Dumbleton JH. Suitability and limitations of carbon fiber reinforced PEEK composites as bearing surfaces for total joint replacements. Wear. 1999, 225-229: 724-727. 12. Kurtz SM and Devine JN. PEEK biomaterials in trauma, orthopedic, and spinal implants. Biomaterials. 2007, 28(32): 4845-4869.

8 13. Robotti P and Zappini G. Thermal Plasma Spray Deposition of Titanium and Hydroxyapatite on Polyaryletheretherketone Implants in PEEK Biomaterials Handbook, edited by S.M. Kurtz. Elsevier Inc; 2012:119-143. 14. Manaster BJ. Total hip arthroplasty: radiographic evaluation. Radiographics. 1996, 16(3): 645-660. 15. Weissman BN. Imaging of total hip replacement. Radiology. 1997, 202(3): 611-623. 16. Ostlere S and Soin S. Imaging of prosthetic joints. Imaging. 2014,15(4): 270–285.

17. Miller TT. Imaging of hip arthroplasty. Eur J Radiol. 2012, 81(12): 3802-3812. 18. Toth JM, Wang M, Estes BT, Scifert JL, Seim HB et al. Polyetheretherketone as a biomaterial for spinal applications. Biomaterials. 2006, 27(3):324-334.

19. Kurtz SM and Nevelos J. Arthroplasty Bearing Surfaces. In PEEK Biomaterials Handbook edited by Kurtz SM. Elsevier Inc; 2012:261-276.

20. Gu S, Rasimick BJ, Deutsch AS, Musikant BL. Radiopacity of dental materials using a digital X-ray system. Dent Mater. 2006, 22: 765-770.

21. Rasimick BJ, Shah RP, Musikant BL, Deutsch AS. Radiopacity of endodontic materials on film and a digital sensor. J Endod. 2007, 33:1098-101.
