The effect of physical loading on calcaneus quantitative ultrasound measurement: a cross-section study
© Liu et al.; licensee BioMed Central Ltd. 2012
Received: 22 May 2011
Accepted: 17 April 2012
Published: 14 May 2012
Physical loading leads to a deformation of bone microstructure and may influence quantitative ultrasound (QUS) parameters. This study aims at evaluating the effect of physical loading on bone QUS measurement, and further, on the potential of diagnosing osteoporosis using QUS method under physical loading condition.
16 healthy young females (control group) and 45 postmenopausal women (divided into 3 groups according to the years since menopause (YSM)) were studied. QUS parameters were measured at calcaneus under self-weight loading (standing) and no loading (sitting) conditions. Weight-normalized QUS parameter (QUS parameter measured under loading condition divided by the weight of the subject) was proposed to evaluate the influence of loading. T-test, One-Way analysis of variance (one way ANOVA) and receiver operating characteristic (ROC) analysis were applied for analysis.
In QUS parameters, mainly normalized broadband ultrasound attenuation (nBUA), measured with loading significantly differed from those measured without loading (p < 0.05). The relative changes of weight-normalized QUS parameters on postmenopausal women with respect to premenopausal women under loading condition were larger than those on traditional QUS parameters measured without loading. In ROC analysis, weight-normalized QUS parameters showed their stronger discriminatory ability for menopause.
Physical loading substantially influenced bone QUS measurement (mainly nBUA). Weight-normalized QUS parameters can discriminate menopause more effectively. By considering the high relationship between menopause and osteoporosis, an inference was drawn that adding physical loading during measurement may be a probable way to improve the QUS based osteoporosis diagnosis.
KeywordsQuantitative ultrasound Osteoporosis Physical loading Menopause
It has been a long time since osteoporosis was recognized as a major public health problem, with enormous social and economic impact . Currently, the best validated method for osteoporosis diagnosis is dual X-ray absorptiometry (DXA) [2, 3]. However, this method is expensive and requires skilled radiographers. At the same time, the bone mineral density (BMD) measured by DXA only provides information on bone mass but contains little information on the microstructure and mechanical properties of the bone which are also important factors of osteoporosis [4, 5]. In recent years, quantitative ultrasound (QUS) has been recognized as a novel technique for osteoporosis diagnosis . Although the sensitivity of QUS is not enough for a clinical osteoporosis diagnosis , it has other advantages such as low cost, with no radiation, and most importantly, providing comprehensive information on both bone mass and the microstructure of the bone [8–10]. As a result, QUS has been recommended as an effective screening technique for digging out osteoporosis suspects from large populations .
In previous studies, QUS measurement was normally performed on bone specimens without physical loading (sitting) [7–12]. However, in situ human bones, especially those composing the axial and appendicular skeleton, are subjected to repetitive, cyclic loading during routine daily activities. Experimental results showed that the microstructure of the bone deformed and mechanical properties of the bone changed significantly under physical loading [13, 14]. On the other hand, the propagation of ultrasound wave is influenced not only by bone mass but also by bone microstructure and material properties [15–18]. As a result, it is possible that the physical loading exerted on bone will not change BMD (due to the mineral content of the bone is not changed) but the QUS parameters (due to the microstructure of the bone deformed and mechanical properties of the bone changed under loading condition). In addition, these changes of QUS parameters should be more on bones with osteoporosis, since bones with osteoporosis are easier to distort than normal bones under the same loading condition. Then we proposed the hypothesis that more distortions on the microstructures of osteoporosis bones under loading condition may help to discriminate the osteoporosis bones more effectively from normal bones using bone QUS measurement (i.e. adding loading may help improve the sensitivity of QUS based osteoporosis diagnosis). Since low sensitivity (i.e. overlapping of QUS parameters measured from osteoporosis patients and normal people) is currently the most significant problem in QUS based osteoporosis diagnosis , if this hypothesis is valid, the findings will help to find a possible way to overcome this significant shortcoming in bone QUS osteoporosis diagnosis. Based on our hypothesis, an experiment was conducted in this preliminary study in order to verify: 1) whether bone QUS parameters measured under loading condition differ from those measured without loading; 2) if the answer to 1) is “yes”, is it possible that exerting loading during ultrasound measurement can provide more useful information to osteoporosis diagnosis, or on the other word, to potentially improve the osteoporosis diagnosis based on QUS?
In this study, QUS measurements were conducted at calcaneus on pre and postmenopausal women under both loading (standing) and no loading (sitting) conditions. “Standing” means that the subject’s self-weight is used as loading exerted on calcaneus during measurements. By considering the consistent and high relationship between menopause and osteoporosis [4, 19, 20], the postmenopausal women were considered as approximations to osteoporosis or osteopenia. To our best knowledge, this is the first study which investigates the effect of loading on the bone QUS measurement.
In total, 61 females were enrolled, including 16 healthy females aged below 30 years and 45 postmenopausal women aged from 50 to 77 years. All subjects were recruited from the same region of the country and were divided into four groups, one control group (group 0, n = 16, age: 23.4 ± 1.05 years, weight: 50.1 ± 4.9 kg) and three testing groups (postmenopausal women) according to the years since menopause (YSM): 1 ≤ YSM ≤ 10 for group1 (n = 12, age: 56.2 ± 2.8 years, weight: 63.1 ± 8.7 kg, YSM: 5.0 ± 2.0 years), 11 ≤ YSM ≤ 20 for group 2 (n = 18, age: 63.4 ± 4.2 years, weight: 62.8 ± 9.7 kg, YSM: 14.5 ± 2.9 years) and 21 ≤ YSM ≤ 30 for group 3 (n = 15, age: 73.2 ± 2.9 years, weight: 59.7 ± 8.9 kg, YSM: 25.7 ± 2.8 years). Informed consents were obtained from all participants before the study. This study was approved by the Beihang university ethics committee.
Calculation of QUS parameters
The precisions characteristics were estimated from triplicate measurements of 30 subjects. Coefficient of variation (CV %) were 1.53%, 0.24% and 1.87% for nBUA, SOS and STI, respectively. Absolute precision error (root mean square standard deviation: RMS-SD) were 0.980 dB·MHz-1·cm-1, 3.807 m·s-1 and 1.306 for nBUA, SOS and STI, respectively.
where ‘ΔX’ indicates the relative change of parameter ‘X’; ‘’ means the mean value of ‘X’ in the control group; and ‘’ indicates the mean value of ‘X’ in the i th group (i = 1, 2, 3).
where ‘T’ indicates the T score of parameter ‘X’, ‘’ means the mean value of ‘X’ in the control group (a young healthy population) and ‘’ indicates the standard deviation of ‘’. Please note that T scores calculated here are different from Z scores which are calculated also by equation (6) but values of the patient's age-matched population are used as the reference values instead. Then, a receiver operating characteristic (ROC) analysis was performed to investigate the discriminatory ability of QUS parameters for menopause.
QUS parameters measured among different subject groups
Group 0 (control)
1596.84 ± 33.16
1570.24 ± 36.98a
1566.80 ± 32.60a
1553.51 ± 18.83a
19.66 ± 4.90
15.22 ± 3.18b
13.95 ± 5.87b
11.72 ± 3.19b,c
86.08 ± 19.44
70.12 ± 16.69a
65.32 ± 22.19a
55.65 ± 14.24a
1596.67 ± 30.18
1574.07 ± 35.83a
1568.63 ± 31.77a
1558.36 ± 15.60a
18.65 ± 4.18
14.57 ± 2.92b
12.97 ± 5.45b
11.25 ± 3.07b,c
83.21 ± 16.66
69.14 ± 15.96b
68.94 ± 14.76b
52.75 ± 9.66b,c
The results of weight-normalized QUS parameters
32.13 ± 3.10
0.39 ± 0.10
1.72 ± 0.38
25.31 ± 3.43a
0.25 ± 0.06a
1.13 ± 0.30a
25.48 ± 3.82a
0.23 ± 0.11a
1.07 ± 0.42a
26.58 ± 3.88a
0.20 ± 0.05a,b
0.94 ± 0.22a
The differences between QUS parameters obtained with and without loading ( P values of Paired-Samples T-test)
wl_nBUA Vs. l_nBUA
p = 0.002a
p = 0.007a
p = 0.001a
p = 0.009a
wl_SOS Vs. l_SOS
p = 0.899
p = 0.024a
p = 0.210
p < 0.001a
wl_STI Vs. l_STI
p = 0.017a
p = 0.270
p = 0.296
p = 0.126
Relative changes of mean QUS parameters in postmenopausal groups with respect to control group (%)
relative changes of mean speed of sound
relative changes of mean normalized broadband ultrasound attenuation
relative changes of mean stiffness index
The discriminatory ability of QUS parameters for menopause
Asymptotic 95% CI
0.749 - 0.912
0.733 - 0.903
0.868 - 0.975
0.763 - 0.926
0.761 - 0.926
0.854 - 0.956
0.719 - 0.899
0.710 - 0.887
0.856 - 0.971
The main finding of our study is the self-weight loading mainly influences the ultrasound attenuation in calcaneus QUS measurement. And the loading-induced changes on QUS parameters were more on postmenopausal women than those on premenopausal women. The results of this study indicated that calcaneus QUS measurement performed with loading may be a probable way to improve the QUS based osteoporosis diagnosis.
It is well-known that the BMD measured by DXA is the golden standard for osteoporosis diagnosis. However, in this study we did not use the BMD as the reference indicator of osteoporosis. The reason why we did so are two: firstly, we consider the radiation of DXA measurement may do harm to the subjects; secondly, this is a preliminary study in which we just want to address the potential but not to draw the conclusion whether adding a physical loading can improve the QUS based osteoporosis diagnosis. Therefore, considering the high and direct correlation between menopause and osteoporosis, we used the postmenopausal women as approximations of osteoporosis or osteopenia in this study when DXA measurement was not performed. Moreover, we understand that further tests using DXA measurement as diagnosis reference must be performed in proceeding study to conclude whether adding a physical loading can improve the QUS based osteoporosis diagnosis.
Since trabecular bone is an inhomogeneous porous medium, the interaction between physical loading and bone and the interaction between ultrasound and bone are highly complex phenomenon. Modelling ultrasonic propagation through trabecular bone tissue using porous media theories showed that the sound velocity and attenuation in trabecular bone depend on the frequency of the ultrasound, the elastic properties of the constituting materials, porosity, permeability, tortuosity, and effective stress . Previous studies also suggested that QUS parameters were highly associated with bone microstructure and mechanical properties [15–18]. While a loading is exerted on bone, the bone trabecula buckle and bend, and a strain of the bone takes place. The deformation of the bone may change the microstructure and some other properties of the bone (e.g. the porosity and permeability), resulting in the changes of reflection and scattering of the ultrasound signal transmitting through the bone, thus the ultrasound property parameters change accordingly.
Paired sample T-test was used to determine whether there were significant differences between pairs of values measured on the same subject under two different loading conditions. The testing results (shown in Table 3) indicated that QUS parameters, mainly nBUA, significantly changed when loading was exerted on calcaneus. The nBUAs obtained with and without loading were different in all subject groups. We presumed that the deformation of the bone microstructure under self-weight loading is the main reason for these differences. As pointed out by Haiat et al , nBUA showed strong correlations with the ratio of bone volume to tissue volume (BV/TV) that 20% change of BV/TV resulted in a change of nBUA by 45 dB·MHz-1·cm-1. In our study, the propagation path of the ultrasound beam can be regarded as a cylinder of bone tissue. Since the diameter of the transducers (i.e. the diameter of the cross section of the cylinder) and the width of the calcaneus (i.e. the height of the cylinder) are constants, the volume of the cylinder is a constant (i.e. the tissue volume (TV) is a constant) either with or without loading. The calcaneus trabecula however buckle and bend when loading is applied, and these deformations lead to the strain of bone. The strain induced by loading will reduce the space among trabecula, compress more bone tissue into the ultrasound path, increase the value of BV, and then increase the value of BV/TV. And the ultrasound attenuation increased correspondingly.
It was found in our results that the sensitivities of SOS and STI to loading were lower than that of nBUA. There were only statistically significant differences between SOS measured with and without loading in group 1 and 3. Different behaviours of SOS and nBUA to loading may probably result from the fact that nBUA is mainly influenced by the microstructure of the bone while SOS is mainly determined by bone density [7–10]. Self-weight loading causes deformations on the bone (deformation of microstructure); however, it does not alter the density and elasticity of the bone. In addition, as a combination of BUA and SOS, STI measured with loading only differed from that measured without loading in the control group. One possible reason is that STI was calculated from an empirical equation (equation (3)) based on the experimental results measured without loading and this equation may be no longer suitable for calculating STI under loading condition.
Our results showed that some QUS parameters, mainly nBUA, substantially changed if loading was exerted on bone during measurement. Then, it should be validated that whether the changes of QUS parameters induced by loading influence the QUS based osteoporosis diagnosis or not. The experimental results reported by Wright et al  may provide some hints to us. According to their study, under the same loading stress, the strain of healthy bone was much lower than that of the decalcified and deproteinized bones. Since the calcium and protein in the osteoporotic bone are lower than those in the healthy bone, the strain of healthy bone should be lower than that of the osteoporotic bone under the same loading condition. That results in smaller deformations on the microstructure of the healthy bone than that of the osteoporotic bone. The resultant changes of QUS parameters of the healthy bone should be smaller than those of the osteoporotic bone. As a result, the differences on QUS parameters between healthy and osteoporotic bones should be enlarged when the QUS measurement is performed with the same loading exerted on the bone.
The analyses above can be well validated by the data shown in Table 4. Although the relative changes of QUS parameters in postmenopausal women with respect to healthy women under loading condition were similar to those measured without loading, the relative changes of weight-normalized QUS parameters (i.e. with loading and all subjects subject to approximately the same stress -1kg stress) were much higher than those of the QUS parameters obtained without loading. These results indicated that the differences between pre and postmenopausal women were larger when weight-normalized QUS parameters were used.
In ROC analysis, the AUCs of QUS parameters measured with loading were not higher (even a few lower) than those measured without loading. However, after corresponding QUS parameters being normalized by weight, the AUCs increased significantly (higher than 0.9) from their counterparts obtained from QUS parameters measured without loading. This meant that only the QUS parameters measured under the same amount of physical loading showed stronger discriminatory ability for menopause in this study. On another word, weight-normalized QUS parameters eliminated the variations of weight induced loadings which exerted on different subjects, and revealed the true effect of physical loading on ultrasound measurement. The results of ROC analysis suggested that the discriminatory ability of weight-normalized QUS parameters for menopause was stronger than that of traditional QUS parameters.
To our best knowledge, this is the first study that investigated the effect of loading on the bone QUS measurement and further proposed the weight-normalized QUS parameters. The weight-normalized QUS parameters were proposed by exerting the physiological loading on traditional QUS parameters. In this study, weight-normalized QUS parameters were significantly lower in postmenopausal groups than those in the control group and the relative changes of the weight-normalized QUS parameters after menopause were much higher than those measured without loading. It is worth noting that there is more weight in subjects in postmenopausal groups than those in the premenopausal group in this study. However, this fact cannot be the main reason for the results above, since previous studies [26, 27] have shown that QUS parameters, including SOS, nBUA and STI, were in positive correlation to weight. It means that more weight companioned with higher values of QUS parameters. Meanwhile, the mean weight of group 3 was the lowest in postmenopausal groups, but the corresponding mean value of weight-normalized nBUA and weight-normalized STI were the lowest and the relative changes of weight-normalized nBUA and weight-normalized STI were the highest in postmenopausal groups. So the lower values and the higher relative changes of weight-normalized QUS parameters in postmenopausal groups cannot be simply ascribed to more weight in subjects in postmenopausal groups.
In this study, the effect of loading was adopted as an initial attempt to bone QUS measurement. Although there are some limitations on these parameters, our results suggested that weigh-normalized QUS parameters could discriminate pre and postmenopausal women more effectively. And by considering the high relationship between menopause and osteoporosis, weight-normalized QUS parameters showed its potential advantages on osteoporosis diagnosis and further investigations about how much stress exerted on the bones is suitable for this method and how to exert the same stress on different subjects are needed.
QUS parameters (mainly nBUA) measured on the calcaneus significantly changed after loading was exerted during QUS measurement. Between pre and postmenopausal women, the differences on weight-normalized QUS parameters were larger than those on traditional QUS parameters. Weight-normalized QUS parameters measured under loading condition can discriminate postmenopausal women from premenopausal women more effectively. This study reveals that loading is an important factor which mainly influences bone ultrasound attenuation, and furthermore, we infer that adding physical loading during QUS measurement may be a probable way to improve the QUS based osteoporosis diagnosis. Further studies will be conducted next to reveal the mechanism of physical loading on QUS measurement.
This study was supported by the National Key Technology R & D Program 2009BAI79B03, and the National Nature Science Foundation of China 10925208 and 61101008.
- Blanchard F: Report on osteoporosis in the european community: building strong bones and preventing fractures-action for prevention. 1998, European Communities, In. BrusselsGoogle Scholar
- Kanis JA, Gluer CC: An update on the diagnosis and assessment of osteoporosis with densitometry. Committee of Scientific Advisors, International Osteoporosis Foundation. Osteoporos Int. 2000, 11 (3): 192-202. 10.1007/s001980050281.View ArticlePubMedGoogle Scholar
- Johnell O, Kanis JA, Oden A, Johansson H, De Laet C, Delmas P, Eisman JA, Fujiwara S, Kroger H, Mellstrom D: Predictive value of BMD for hip and other fractures. J Bone Miner Res. 2005, 20 (7): 1185-1194. 10.1359/JBMR.050304.View ArticlePubMedGoogle Scholar
- Tauchmanova L, Nuzzo V, Del PA, Fonderico F, Esposito-Del PA, Padulla S, Rossi A, Bifulco G, Lupoli G, Lombardi G: Reduced bone mass detected by bone quantitative ultrasonometry and DEXA in pre and postmenopausal women with endogenous subclinical hyperthyroidism. Maturitas. 2004, 48 (3): 299-306. 10.1016/j.maturitas.2004.02.017.View ArticlePubMedGoogle Scholar
- Goldstein SA, Goulet R, McCubbrey D: Measurement and significance of three-dimensional architecture to the mechanical integrity of trabecular bone. Calcif Tissue Int. 1993, 53 (Suppl 1): S127-S132. S132-S133View ArticlePubMedGoogle Scholar
- Resch H, Pietschmann P, Bernecker P, Krexner E, Willvonseder R: Broadband ultrasound attenuation: a new diagnostic method in osteoporosis. AJR Am J Roentgenol. 1990, 155 (4): 825-828.View ArticlePubMedGoogle Scholar
- Gluer CC: Quantitative ultrasound techniques for the assessment of osteoporosis: expert agreement on current status. The international quantitative ultrasound consensus group. J Bone Miner Res. 1997, 12 (8): 1280-1288. 10.1359/jbmr.1918.104.22.1680.View ArticlePubMedGoogle Scholar
- Hans D, Wu C, Njeh CF, Zhao S, Augat P, Newitt D, Link T, Lu Y, Majumdar S, Genant HK: Ultrasound velocity of trabecular cubes reflects mainly bone density and elasticity. Calcif Tissue Int. 1999, 64 (1): 18-23. 10.1007/s002239900572.View ArticlePubMedGoogle Scholar
- Nicholson PH, Muller R, Cheng XG, Ruegsegger P, Van Der Perre G, Dequeker J, Boonen S: Quantitative ultrasound and trabecular architecture in the human calcaneus. J Bone Miner Res. 2001, 16 (10): 1886-1892. 10.1359/jbmr.2001.16.10.1886.View ArticlePubMedGoogle Scholar
- Muller M, Mitton D, Moilanen P, Bousson V, Talmant M, Laugier P: Prediction of bone mechanical properties using QUS and pQCT: Study of the human distal radius. Med Eng Phys. 2008, 30 (6): 761-767. 10.1016/j.medengphy.2007.08.006.View ArticlePubMedGoogle Scholar
- Barkmann R, Laugier P, Moser U, Dencks S, Padilla F, Haiat G, Heller M, Gluer CC: A method for the estimation of femoral bone mineral density from variables of ultrasound transmission through the human femur. Bone. 2007, 40 (1): 37-44. 10.1016/j.bone.2006.07.010.View ArticlePubMedGoogle Scholar
- Trimpou P, Bosaeus I, Bengtsson B, Landin-Wilhelmsen K: High correlation between quantitative ultrasound and DXA during 7 years of follow-up. Eur J Radiol. 2010, 73: 360-364. 10.1016/j.ejrad.2008.11.024.View ArticlePubMedGoogle Scholar
- Qin YX, Lam H, Zhang M, Orzechowski L: Dynamic skeletal muscle contraction induced bone fluid pressure and its role in adaptation. J Biomech. 2007, 40 (S2): S37-View ArticleGoogle Scholar
- Muller R, Gerber SC, Hayes WC: Micro-compression: a novel technique for the nondestructive assessment of local bone failure. Technol Health Care. 1998, 6 (5–6): 433-444.PubMedGoogle Scholar
- Biot MA: Theory of propagation of elastic waves in a fluid-saturated porous solid. II. Higher frequency range. J Acoust Soc Am. 1956, 28 (2): 179-191. 10.1121/1.1908241.View ArticleGoogle Scholar
- Barkmann R, Heller M, Glüer CC: Methoden der in vivo-ultraschall-messtechnik am skelett: grundlagen und technische realisierung. J Miner Stoffwechs. 1999, 6 (4): 22-27.Google Scholar
- Njeh CF, Fuerst T, Diessel E, Genant HK: Is quantitative ultrasound dependent on bone structure? A reflection. Osteoporos Int. 2001, 12 (1): 1-15.PubMedGoogle Scholar
- Nicholson PH, Cheng XG, Lowet G, Boonen S, Davie MW, Dequeker J, Van der Perre G: Structural and material mechanical properties of human vertebral cancellous bone. Med Eng Phys. 1997, 19 (8): 729-737. 10.1016/S1350-4533(97)00030-1.View ArticlePubMedGoogle Scholar
- Vu T, Khan NC, Lam NT, Mai LB, DucSon N, Nhung BT, Nakamori M, Kunii D, Sakai T, Yamamoto S: Determining the prevalence of osteoporosis and related factors using quantitative ultrasound in Vietnamese adult women. Am J Epidemiol. 2005, 161 (9): 824-830. 10.1093/aje/kwi105.View ArticlePubMedGoogle Scholar
- Falgarone G, Porcher R, Duche A, Kolta S, Dougados M, Roux C: Discrimination of osteoporotic patients with quantitative ultrasound using imaging or non-imaging device. Joint Bone Spine. 2004, 71 (5): 419-423. 10.1016/j.jbspin.2003.09.011.View ArticlePubMedGoogle Scholar
- Langton CM, Palmer SB, Porter RW: The measurement of broadband ultrasonic attenuation in cancellous bone. ARCHIVE: Engineering in Medicine. 1984, 13 (2): 89-91. 1971–1988 (vols 1–17)Google Scholar
- Serpe LJ, Rho JY: Broadband ultrasound attenuation value dependence on bone width in vitro. Phys Med Biol. 1996, 41: 197-10.1088/0031-9155/41/1/014.View ArticlePubMedGoogle Scholar
- Langton CM, Njeh CF: The measurement of broadband ultrasonic attenuation in cancellous bone-a review of the science and technology. Ultrasonics, Ferroelectrics and Frequency Control, IEEE Transactions on. 2008, 55 (7): 1546-1554.View ArticleGoogle Scholar
- Haiat G, Padilla F, Barkmann R, Gluer CC, Laugier P: Numerical simulation of the dependence of quantitative ultrasonic parameters on trabecular bone microarchitecture and elastic constants. Ultrasonics. 2006, 44: e289-e294.View ArticlePubMedGoogle Scholar
- Wright TM, Vosburgh F, Burstein AH: Permanent deformation of compact bone monitored by acoustic emission. J Biomech. 1981, 14 (6): 405-409. 10.1016/0021-9290(81)90058-0.View ArticlePubMedGoogle Scholar
- Drozdzowska B: Quantitative ultrasound measurements at the calcaneus in natural and surgically induced menopause. Maturitas. 2006, 53 (1): 107-113. 10.1016/j.maturitas.2005.03.005.View ArticlePubMedGoogle Scholar
- Adami S, Giannini S, Giorgino R, Isaia G, Maggi S, Sinigaglia L, Filipponi P, Crepaldi G, Di Munno O: The effect of age, weight, and lifestyle factors on calcaneal quantitative ultrasound: the ESOPO study. Osteoporos Int. 2003, 14 (3): 198-207.PubMedGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2474/13/70/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.