BJR
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
First published online September 17, 2007
British Journal of Radiology (2007) 80, 919-925
© 2007 British Institute of Radiology
doi: 10.1259/bjr/33589854
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Economos, C D
Right arrow Articles by Naumova, E N
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Economos, C D
Right arrow Articles by Naumova, E N

Full paper

Precision of Lunar Achilles+ bone quality measurements: time dependency and multiple machine use in field studies

C D Economos, PhD 1 J M Sacheck, PhD 1 W Wacker, BS 2 K Shea, MS 1 and E N Naumova, PhD 3

1 Tufts University, Gerald J and Dorothy R Friedman School of Nutrition Science and Policy, 150 Harrison Avenue, Boston, MA 02111, 2 GE Healthcare, 726 Heartland Trail, Madison, WI 53717, 3 Division of Public Health and Family Medicine, Tufts University School of Medicine, 136 Harrison Avenue, Boston, MA 02111, USA

Correspondence: Jennifer Sacheck, Gerald J and Dorothy R Friedman School of Nutrition Science and Policy, Tufts University, 150 Harrison Avenue, Boston, MA 02111, USA. E-mail: jennifer.sacheck{at}tufts.edu


   Abstract
Top
 Abstract
Introduction
 Methods
 Results
 Discussion
 References
 
Qualitative ultrasound (QUS) is a portable, safe and relatively inexpensive technique used to obtain information on bone mineral quality in adults and children. QUS measures bone stiffness index (SI) through the incorporation of speed of sound (SOS) and broadband ultrasound attenuation (BUA). QUS technology may prove to be extremely useful in field research where more than one machine is used over different periods of time. 13 adults (27.6±4.6 years old) were recruited to determine the internal stability of two Lunar Achilles+ QUS machines (Lunar1, Lunar2), as well as the repeatability in bone stiffness measures between the two machines over time. Triplicate measurements of the calcaneus were taken within the same day (n = 258) and at 1 week (n = 120), 6 months (n = 54) and 1 year (n = 18) apart to determine the time-dependent repeatability. Using paired t-tests and separate mixed effects models, there were no differences reported in SI, SOS or BUA values within one machine, or between two machines over these short- and long-term time-frames. These results indicate that QUS machines are internally consistent and different machines may be used over time to provide reliable measurements of changes in bone quality.


   Introduction
Top
 Abstract
 Introduction
Methods
 Results
 Discussion
 References
 
Although the most valid method for measuring bone density is dual X-ray absorptiometry (DXA), there have been a number of devices developed recently to measure skeletal status. Quantitative ultrasound (QUS) provides a safe, non-invasive means to obtain information on bone quality (defined scientifically as "acoustic bone quality" that, for purposes of simplicity in this paper, we will refer to as "bone quality") in adults [1–3] and children [4, 5] by measuring a peripheral site (such as heel or wrist) without using ionizing radiation. In addition, QUS provides a more manageable means of collecting bone mineral measurements in field studies, as it is less costly than DXA, quick and easy to use, and fairly transportable. Importantly, QUS scores have been shown to be predictive of fracture risk in older adults [6, 7] and offer diagnostic sensitivity [8]. Risk factors usually associated with DXA measures in elderly women (e.g. physical activity, calcium intake, history of cigarette smoking) are also associated with QUS measures in pre-menopausal women [9]. In addition, QUS has been used to measure bone quality in paediatric populations [5, 10–12] and has been found to be a valuable screening tool for young patients at risk for osteopenia [13, 14]. Thus, QUS measurements can help to identify both young and older individuals at risk for osteoporosis.

The Lunar Achilles+ (GE Healthcare, Milwaukee, WI) is a QUS calcaneal device that is small, portable, inexpensive, FDA-approved, and offers excellent short-term precision [6, 15–17]. Lunar Achilles+ measurements in children have provided reproducible results [4, 11] and have a similar precision to those found in adults [4, 14]. The coefficient of variation for in vivo calcaneus measurement by QUS is 2.2–2.7% in adults [3, 6, 18] and 1.9–3.5% in children [4, 11, 14]; precision as given by the manufacturer is 1.7%. QUS uses sound waves to assess bone quality by two parameters: (i) ultrasound velocity or speed of sound (SOS (m s–1)), which exploits the speed of the ultrasound wave to assess elasticity and bone density [19]; and (ii) broadband ultrasound attenuation (BUA (dB MHz–1)), which exploits the frequency dependence of ultrasound attenuation to determine bone architecture and density [20]. An additional parameter, the stiffness index (SI), is calculated as a percentage of values obtained by the manufacturer in a young adult population by a linear combination of SOS and BUA (SI = (0.67 x BUA) + (0.28 x SOS) – 420). The SI was developed to improve the standardized coefficient of variation of SOS or BUA alone and also to compensate for water bath temperature variation [21]. These bone ultrasound property measures are highly correlated with DXA measurements in children [14], adolescents [22] and in young [23] and older adults [2, 3, 6].

In large-cohort field-based research, it may be necessary to use more than one QUS machine, along with more than one tester, in order to collect measurements in a timely and efficient manner. The goal of this examination was to establish short- and long-term repeatability in bone stiffness measures using two Lunar Achilles+ machines by examining the SI measured within 1 day, and subsequently at 1 week, 6 months and 1 year, and then to compare the results for each machine. This analysis was designed to answer the following questions:

  1. When measuring one individual, what is the internal stability of each Lunar machine at short- and long-term intervals?
  2. Are the readings or the variability of readings on one machine different from on the other?
  3. Do measurements performed by different testers on one individual yield different results?

We hypothesized that each machine would yield repeatable results over time that are highly comparable between the two machines and amongst testers, and that the variability would be less than the expected change with growth over time. These data were collected to inform researchers about the utilization of QUS machines in both short- and long-term field studies, and to demonstrate their utility in assessment of bone quality and status.


   Methods
Top
 Abstract
 Introduction
 Methods
Results
 Discussion
 References
 
Study design
This analysis was designed to examine the internal and temporal stability in the performance of two Lunar Achilles+ machines (Lunar1 and Lunar2). The examination was carried out over a 1-year period with four measurement time periods (same day, 1 week, 6 months and 1 year). There were a total of 13 healthy Caucasian participants (median age 27.0±5.7 years for females, n = 9; and 27.5+3.9 years for males, n = 4). This age range was chosen under the assumption that bone density should be at a plateau compared with younger (<20 years) and older (>35 years) individuals, and therefore ideal for investigating machine stability over time. QUS was performed using the Lunar Achilles+ on the left heel (calcaneus). Both machines were calibrated and quality assurance run on a regular basis according to the manufacturer's recommended instructions.

In this study design, human participants were used to determine internal temporal stability of the Lunar machines without the addition of phantoms for two important reasons. Firstly, the manufacturer (GE Healthcare) has found that, unlike the case for X-ray densitometers, phantoms are not particularly good for measuring the stability of bone ultrasonometers (personal communication; R Morris and W Wacker, GE Healthcare, January 2006). Secondly, use of the Achilles+ phantom for monitoring stability in large-scale clinical studies has been found to be unsuitable for detecting long-term drift [24]. Thus, we took measurements on human participants after applying alcohol spray to the heel as recommended by the manufacturer. Measurements were taken in triplicate (with each measurement consisting of a BUA, SOS and SI score) on each of the two machines on the same day (six measurements in total per day), with repositioning of the foot between measurements. The triplicate measurement design in each set contains the same information as would be obtained in a classical precision study, and these data are presented as coefficients of variation (CVs) in the Results section. The order of machine assignments was randomized on each measurement day. The Institutional Review Board stated that this was a quality assurance analysis and therefore not an IRB-approved research study. Accordingly, all subjects signed a standard waiver of liability approved by the Jean Mayer USDA Human Nutrition Research Center on Aging (HNRCA) at Tufts University prior to participation.

Subjects were employee volunteers from the HNRCA and trained by a research associate on operating the Lunar Achilles+ using a standardized protocol. Under the supervision of this staff member, each subject read his/her own SI, SOS and BUA measurements from the Achilles+ software (version 3.04) and recorded the measurement on a datasheet. Each subject's measurements were also automatically saved into the Lunar Achilles+ database interfaced with a computer. The research associate set up the machine and supporting materials at set time intervals so that each subject could conduct this process on 2 days exactly 1 week apart. This process was then repeated again at 6 months and 1 year following the start of the study for a total possible of six testing days (Figure 1Go). This design yielded 43 pairs of triplicate measurements to compare same-day (within one day; n = 258, 13 subjects); 20 pairs to compare short-term (1 week; n = 120, 11 subjects); 9 pairs to compare intermediate (6 months; n = 54, 6 subjects); and 3 pairs to compare long-term (1 year; n = 18, 3 subjects) differences.


Figure 1
View larger version (15K):
[in this window]
[in a new window]

  		Figure 1. Schematic diagram of the study design. Two Lunar Achilles+ ultrasound machines were assessed over different time periods. Six same-day trials (D1 and D2s) took place over the period of 1 year. Each trial day consisted of six consecutive measurements: three on Lunar1 (L1) and three on Lunar2 (L2). These within-day measurements took place at the beginning or end of a week, such that there were three week-long trials. The three testing weeks were scheduled at 6 month intervals yielding two 6-month examination periods. Finally, these two 6-month time periods constituted one 1-year test period. Subjects were scheduled for testing on as many D1s and D2s as possible throughout this 1-year testing period. n = number of measurements during each trial period. Trial days are shaded in grey.

 
Statistical analysis
Statistical analyses were performed using S-plus statistical software (Insightful Inc., Seattle, WA). A paired t-test was used to determine the differences between measured values of SI, SOS and BUA that were obtained from Lunar1 and Lunar2. Pearson's correlation coefficients were used to determine a linear association between the measurements of the two machines. Sex-related differences in QUS scores were calculated by an independent samples t-test. Same-day, 1 week and 6 month CVs of SI, SOS and BUA were calculated by dividing the standard deviation for triplicate measurements by the mean of the measurements for the given time period, as recommended by the International Society for Clinical Densitometry. Simple linear regressions of SI, SOS and BUA values between Lunar1 and Lunar2 were used for verification and explanation of outliers. To visualize the data, box-plots and Bland–Altman plots were created [25].

To establish repeatability between the two machines within the same day, and over 1 week, 6 months and 1 year, the differences in measures of SI, SOS and BUA were examined in separate mixed effects models [26]. We assumed repeated measures of SI, SOS and BUA to be nearly unchanged over the examination period and therefore highly clustered within an individual; however, within each model, adjustments for changes in age were performed. The first model (Model 1) examined the differences in measurements taken on the same day, adjusting for age and sex. The second model (Model 2) examined the differences in measurements performed 1 week apart adjusting for age at second measurement and sex. The third model (Model 3) examined the differences in measurements performed 6 months apart (159–195 days apart) adjusting also for age at second measurement and sex. For the analysis of these data, a linear mixed effects model, with the Gaussian assumption for the outcome, was assumed:


Formula 001

where yijk are the outcome measurements for i (individual), j (measurement) and k (subset); L is an indicator variable for the Lunar machine; age is expressed in full year to two decimal places; and sex is an indicator variable. For each subset, the age and sex variables were added sequentially. beta0 is an intercept of the regression model. The parameter beta1 reflects the systematic differences between two machine performances. The S-plus (Version 4.5/6) command line for the linear mixed effects model was as follows:


Formula 002

A fourth model was constructed to examine the differences between the two machines at 1 year and was adjusted for age but not sex, as there were only three individuals completing this time-span (Model 4). However, the utility of this model is questionable as there were only 18 measurements in its establishment. Thus, in conjunction with this model, we used t-tests to examine the difference in long-term repeatability between Lunar1 and Lunar2.


   Results
Top
 Abstract
 Introduction
 Methods
 Results
Discussion
 References
 
From the 13 subjects that were recruited, we gathered a total of 258 measurements each for SI, SOS and BUA (Table 1Go). As expected, the males had significantly greater SI (p<0.001), SOS (p<0.001) and BUA (p<0.001) measures than the females. All measurements of SI obtained from the 13 subjects are shown in the form of a box-plot (Figure 2Go) where each subject (no.) corresponds to an individual, and the two box-plots within an individual reflect the repeated measures of the SI index for Lunar1 and Lunar2. The range of overall SI values between individuals fell between 74 and 144 (overall average, 102.2±15.5). The variability within an individual was substantially smaller than the overall variability between subjects. In addition, the overall average coefficients of variation for Lunar1 and Lunar2 SI readings were similar, i.e. 0.98 vs 1.26, respectively (p = 0.22), and the overall correlations between Lunar1 and Lunar2 for SI, SOS and BUA were 0.95, 0.96 and 0.92, respectively (p<0.05). When coefficients of variation were separated relative to the time period examined, there were still no significant differences between the two machines in SI, SOS or BUA CV measures (Table 2Go). As the time between measurements increased, the CVs increased for BUA and SI, but not SOS, values.


View this table:
[in this window]
[in a new window]

  		Table 1. Lunar Achilles+ quantitative ultrasound bone quality measures for each subject

 

Figure 2
View larger version (20K):
[in this window]
[in a new window]

  		Figure 2. Box-plots representing all 258 measurements of the stiffness index subdivided by subject and Lunar Achilles+ QUS machine. Dotted box-plots represent measurements from Lunar1; open box-plots represent measurements from Lunar2. Subject no. corresponds to subject no. in Table 1Go.

 

View this table:
[in this window]
[in a new window]

  		Table 2. Time-dependent percentage coefficients of variation in bone quality measures for two Lunar Achilles+ quantitative ultrasound machines over different time periods

 
Figure 3Go shows the plot of differences between estimates of SI sample means for different examination periods and SI average values measured on Lunar1 and Lunar2. Neither the paired t-test (p = 0.30) nor the regression slope (p = 0.23) provided evidence for significant differences between the two machines. In addition, there was no identifiable difference in SI or the magnitude of SI, or between these values, over different time periods. Although one outlier was identified (subject no. 13), it did not significantly affect the analysis and therefore remained in the analysis.


Figure 3
View larger version (24K):
[in this window]
[in a new window]

  		Figure 3. Bland–Altman plot of time-dependent differences between Lunar1 and Lunar2 relative to the average stiffness index. 129 measurements were taken on each machine. Solid and dashed lines represent the mean difference±2 standard deviations (SDs), respectively, of all values (–1.37±3.9).

 
To determine whether both Lunar1 and Lunar2 are consistent in yielding identical measurements over time, mixed effects models were used to identify any differences in performance between the two machines. The results from the models, together with the 95% confidence intervals for the differences in the SI, SOS or BUA values for Model 1 (same-day), Model 2 (1 week), Model 3 (6 months) and Model 4 (1 year), are listed in Table 3Go. The results indicate that the differences in performance between the two machines are minimal, although the estimates of the differences were statistically significant in 7 of the 12 cases. The values of SI and BUA obtained on Lunar2 were less than 0.85 units and 1.1 units higher, respectively, than using Lunar1 at the three time-points examined. For one time-point, the values for SOS obtained using Lunar1 were 1.0 unit higher than using Lunar2. Such differences in performance between the two machines are slight compared with the range of SI, SOS and BUA measures. The low p-values are more indicative of the strength and precision of the statistic for identifying small differences as opposed to meaningful biological and practical differences in machine performance. Furthermore, while checking the models' performances, we identified two potential outliers (subject no. 9 and no. 13) in the same-day repeatability model. After omitting data from these two subjects, the model results indicated no differences between SI, SOS and BUA measures (data not shown).


View this table:
[in this window]
[in a new window]

  		Table 3. Mixed effect models examining the systematic differences between Lunar1 and Lunar2 readings for SI, SOS and BUA

 
Models 1–3 presented in Table 3Go include adjustment for age and sex; the estimates for these models suggest that both machines perform consistently regardless of age or sex. The 1 year models were adjusted for age and incorporated only 18 measurements. However, when t-tests were used to examine 1 year differences, we found no significant difference in repeatability for SI, SOS or BUA measures (p = 0.38, 0.27 and 0.55, respectively) over this longer time-frame.


   Discussion
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
References
 
In large-cohort field-based research examining bone quality, quick, efficient and economically feasible methods are often necessary and may limit the use of DXA to measure bone mineral quality. Here we have examined the Lunar Achilles+ QUS machine to determine both its reliability over short and longer time periods and its consistency between two different machines. Thus, observations were divided amongst subject, machine and the time of measurement. Recordings of SI, SOS and BUA within the same subject group, regardless of when the testing occurred, were highly reproducible on either Lunar Achilles+ machine, which is further demonstrated by the high correlation coefficients produced taking into account the variations in positioning and the time-frame within which this analysis occurred. This suggests that the two machines provide consistent and repeatable measures of SI, SOS and BUA within the same day, or at 1 week, 6 months and 1 year apart. There was no relationship between machine and magnitude of difference from the mean SI for each individual. This would suggest that: (i) the internal stability on a given machine at different time points is high; and (ii) two measurements taken on different machines will not be significantly different at one point in time within a year. Consequently, if two measurements taken from an individual on different machines at different time points are different, then this difference can be attributed to a true biological change in bone. Two different well-maintained machines can therefore be used interchangeably to provide stable measurements of SI, SOS and BUA for at least 1 year.

Bone mineral content is highly dependent on age, and bone quality measures utilizing QUS technology have shown that mean calcaneus SI increases by ~80% between the ages of 7 years and 18 years [4], and then decreases after menopause by ~1.1% per year [3]. Peak bone mass (PBM) is achieved in the late 20s, with 99% of PBM being attained by age 26.2±3.7 years [27]. The repeatability of the measurements, especially over the intermediate and long term, would be difficult to determine in young children and older adults, as these groups would be gaining and losing bone mineral, respectively, at a potentially rapid rate. Therefore, in other studies, only short-term precision has been examined in youths (CV = 3.2%) [4] and in post-menopausal women (CV = 2.2–2.4%) [6, 28]. In addition, the dynamic range of the Lunar Achilles+ is optimized for a low SI, and the precision is compromised for individuals with very dense heels, where the signal can be over-attenuated.

In order to examine the reliability of different machines over time, in the present analysis we chose a population whose bone density theoretically would not be expected to change within 1 year's time. As the median age of our sample is 27 years, these subjects would be expected to be at ~99% of their PBM. These subjects were all healthy young adults and would not be expected to experience any substantial changes in bone mineral density over the course of 1 year. Even though our subjects should have been at a plateau for bone mineral quality, we still controlled for age in the regression models as an additional precaution. Indeed, after controlling for age, the immediate, short-term, intermediate and long-term stability of the measurements, even though statistically significant as a result of our conservative statistics, is not likely to be biologically different. In addition, the CVs for the same-day and 1 week precision of BUA, SOS and SI values were similar to those published by others using Lunar Achilles+ [6, 8, 28].

Owing to the loss of follow-up over the course of the examination periods, mixed effects models were used for the same-day, 1 week and 6 month comparisons, but could not be relied upon to assess for longer term (1 year) repeatability because of sample size constraints. Although the use of a mixed effects model may provide more efficient estimation of the parameters, the results of paired t-tests indicate that there were no differences in repeatability of SI, SOS and BUA over the course of 1 year. This means that differences in individual measurements of SI, SOS and BUA from baseline to 1 year taken on two different machines are not attributable to internal inconsistencies between the two machines. Thus, any measured changes in SI, SOS and BUA over the course of 1 year could indeed be caused by changes in bone quality, even if the measurements were taken on different machines (small changes in bone quality that were beyond the scope of this analysis may also explain the subtle increases in CV over longer trial periods for SI and BUA). Other studies using Lunar Achilles+ have demonstrated that the short-term in vivo precision was only slightly lower in a paediatric population relative to adults [4]. It has also been shown that heel ultrasound correlates well with DXA bone mineral density during times that bone quality is changing, in both children [14, 29] and post-menopausal women [6]. Furthermore, compared with other ultrasound devices, bone quality results from the Achilles+ correlated the highest with bone mineral density measurements from DXA [6, 14].

We have noted that there are two outliers in this analysis. In the case of subject no. 13, the most likely cause is the high calcaneus density. The Achilles dynamic range is optimized for measurements of low calcaneus density, and the reduction of signal at high BUA can result in higher variability in the measurements for such subjects. Different machines vary somewhat in strength of acoustic signal and transducer sensitivity, and so systematic differences between devices are sometimes greater for high-density subjects (personal communication; R Morris and W Wacker, GE Healthcare, January 2006). It is not yet clear why subject no. 9 was an outlier. Studies at GE Healthcare over the years have noted rare individuals with highly variable QUS measurements. For such persons, slight changes in position, or even pressure on the foot, can cause easily measured shifts in QUS values. It is possible that the calcaneus of such individuals has a shape or density distribution which focuses acoustic waves at the detector, and that the focus shifts readily with any change in calcaneus position. However, for both subjects in the current study, these explanations are speculative and the identification remains statistical.

One additional factor that affects the utility of QUS in large long-term studies is the necessity to employ several technicians in order to gather necessary data. Only one other study used multiple testers with the Lunar Achilles+ system, and was done in order to examine possible tester-dependent differences in QUS readings [6]. In that study, only four measurements were performed across 1 week: the first three by one technician and the fourth by the second technician. Therefore, as an additional outcome in the present examination, we tested by using the same two Lunar Achilles+ machines whether multiple QUS test administrators would influence the results taken on the same day within a given individual. Four testers performed QUS measurements on five different subjects in a randomized order within 1 day. The inter-observer CVs were 1.74% for SI, 0.35% for SOS and 1.19% for BUA, all of which fall within the normal range of variability of each of these measures. Our overall precision of SI amongst the four testers was 1.01%, which was similar to the inter-observer CV results from Greenspan et al [6] of 1.72% for QUS and 1.91% from DXA readings. Therefore, different QUS testers should not affect results from the same individual.

One potential limitation of this study was our inability to recruit a larger sample size for the measurement of the stability of readings over long-term intervals, and thus the small sample size severely limits the statistical power of the 1 year comparison. Owing to the small sample size, interpretation of the standard error estimates have to be taken with caution. We felt that it was important to retain all analyses that were part of the original study design and the results are included for completeness, but should be interpreted with caution. An additional limitation is that two QUS instruments may be too few to conclude that all Lunar QUS instruments behave in a similar fashion. However, all studies are limited in their ability to make inferences about an entire population from a sample. Furthermore, this study design incorporates two machines and parallels many field studies in which it is likely that only a few machines would be available for analysis of the study sample. Therefore, this is a model study and we recommend that future studies simultaneously run a quality assurance analysis amongst machines. We have established that the use of QUS machines not only serves as a safe and easy way to assess bone quality but also that the results are promising for looking at changes in bone mineral content over a time period of at least 1 year. In addition, investigators can be confident in using different machines and test administrators in long-term and large-scale intervention studies seeking to examine changes in bone ultrasound properties.


   Acknowledgments
 
This research was funded by the National Institutes of Health, R01-HD37752.

Received for publication June 2, 2006. Revision received December 4, 2006. Accepted for publication April 5, 2007.


   References
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 

  1. Funke M, Kopka L, Vosshenrich R, Fischer U, Ueberschaer A, Oestmann J, Grabbe E. Broadband ultrasound attenuation in the diagnosis of osteoporosis: correlation with osteodensitometry and fracture. Radiology 1995;194:77–81.[Abstract/Free Full Text]
  2. Agren M, Karellas A, Leahey D, Marks S, Baran D. Ultrasound attentuation of the calcaneus: a sensitive and specific discriminator of osteopenia in postmenopausal women. Calcif Tissue Int 1991;48:240–4.[CrossRef][Medline]
  3. Kang C, Speller R. Comparison of ultrasound and dual energy X-ray absorptiometry measurements in the calcaneus. Br J Radiol 1998;71:861–7.[Abstract]
  4. Sawyer A, Moore S, Fielding K, Nix D, Kiratli J, Bachrach L. Calcaneus ultrasound measurements in a convenience sample of healthy youths. J Clin Densitom 2001;4:111–20.[CrossRef][Medline]
  5. Wunsche K, Wunsche B, Fahnrich H, Mentzel H-J, Vogt S, Abendroth K, et al. Ultrasound bone densitometry of the os calcis in children and adolescents. Calcif Tissue Int 2000;67:349–55.[CrossRef][Medline]
  6. Greenspan S, Bouxsein M, Melton M, Kolodny A, Clair J, DeLucca P, et al. Precision and discriminatory ability of calcaneal bone assessment technologies. J Bone Miner Res 1997;12:1303–13.[CrossRef][Medline]
  7. Hans D, Dargent P, Schott A, Sebert J, Cormier C, Kotski P, et al. Ultrasound parameters predict hip fracture independently of hip bone density: The Epidos Prospective Study. J Bone Miner Res 1995;10:S169
  8. Njeh C, Hans D, Li J, Fan B, Fuerst T, He Y, et al. Comparison of six calcaneal quantitative ultrasound devices: Precision and hip fracture discrimination. Osteoporos Int 2000;11:1051–62.[CrossRef][Medline]
  9. Adami S, Giannini S, Giorgino R, Isaia G, Maggi S, Sinigaglia L, et al. Effect of age, weight and lifestyle factors on calcaneal quantitative ultrasound in premenopausal women: the ESOPO study. Calcif Tissue Int 2004;74:317–21.[CrossRef][Medline]
  10. van den Bergh JPW, Noordam C, Ozyilmaz A, Hermus ARMM, Smals AGH, Otten BJ. Calcaneal ultrasound imaging in healthy children and adolescents: relation of the ultrasound parameters BUA and SOS to age, body weight, height, foot dimensions and pubertal stage. Osteoporos Int 2000;11:967–76.[CrossRef][Medline]
  11. Zebaze R, Brooks E, High M, Duty E, Bronson W. Reproducibility of heel ultrasound measurement in prepubescent children. J Ultrasound Med 2003;22:1337–340.[Abstract/Free Full Text]
  12. vandenBergh J, Noordam C, Thijssen J, Otten B, Smals A, Hermus A. Measuring skeletal changes with calcaneal ultrasound imaging in healthy children and adults: the influence of size and location of the region of interest. Osteoporos Int 2001;12:970–9.[CrossRef][Medline]
  13. Fielding K, Nix D, Bachrach L. Comparison of calcaneus ultrasound and dual x-ray absorptiometry in children at risk of osteopenia. J Clin Densitom 2003;6:7–15.[CrossRef][Medline]
  14. Jaworski M, Lebiedowski M, Lorene R, Trempe J. Ultrasound bone measurement in pediatric subjects. Calcif Tissue Int 1995;56:368–71.[CrossRef][Medline]
  15. Machado A, Hannon R, Henry Y, Eastell R. Standardized coefficient of variation for dual energy xray absorptiometry (DXA), quantitative ultrasound (QUS), and markers of bone turnover. J Bone Miner Res 1997;12:S258
  16. Langton CM, Langton DK. Comparison of bone mineral density and quantitative ultrasound of the calcaneus: site-matched correlation and discrimination of axial BMD status. Br J Radiol 2000;73:31–5.[Abstract]
  17. Tromp A, Smit J, Deeg D, Lips P. Quantitative ultrasound measurements of the tibia and calcaneus in comparison with DXA measurements at various skeletal sites. Osteoporos Int 1999;9:230–5.[CrossRef][Medline]
  18. Alenfeld F, Engelke K, Schmidt D, Brezger M, Diessel E, Felsenberg D. Diagnostic agreement of two calcaneal ultrasound devices: the Sahara bone sonometer and the Achilles+. Br J Radiol 2002;75:895–902.[Abstract/Free Full Text]
  19. Turner C, Eich M. Ultrasonic velocity as a predictor of strength in bovine calcaneous bone. Calcif Tissue Int 1991;49:116–19.[Medline]
  20. Langton C. Recent advances in the ultrasonic assessment of bone. In: Bath conference on current research in osteoporosis and bone mineral measurement II. 1992; 44
  21. Njeh C, Hans D, Fuerst T, Gluer C, Genant H, editors. Quantitative ultrasound: assessment of osteoporosis and bone status. London, UK: Martin Dunitz; 1999
  22. Sundberg M, Gardsell P, Johnell O, Ornstein E, Sernbo I. Comparison of quantitative ultrasound measurements in calcaneus with DXA and SXA at other skeletal sites: a population-based study on 280 children aged 11–16 years. Osteoporos Int 1998;8:410–17.[CrossRef][Medline]
  23. Lum CK, Wang M-C, Moore E, Wilson DM, Marcus R, Bachrach LK. A comparison of calcaneus ultrasound and dual X-ray absorptiometry in healthy North American youths and young adults. J Clin Densitom 1999;2:403–11.[CrossRef][Medline]
  24. Hans D, Wacker W, Genton L, Paris E, Le-Floch C, Slosman D. Longitudinal quality control methodology for the quantitative ultrasound Achilles+ in clinical trial settings. Osteoporos Int 2002;13:788–95.[CrossRef][Medline]
  25. Bland J, Altman D. Comparing methods of measurement: why plotting difference against standard method is misleading. Lancet 1995;346:1085–7.[CrossRef][Medline]
  26. Pinheiro J, Bates D. Mixed-effects models in S and S-PLUS. New York: Springer Verlag; 2000
  27. Teegarden D. Peak bone mass in young women. J Bone Miner Res 1995;10:711–15.[Medline]
  28. Krieg M, Cornuz J, Ruffieux C, Sandini L, Buche D, Dambacher M, et al. Comparison of three bone ultrasounds for the discrimination of subjects with and without osteoporotic fractures among 7562 elderly women. J Bone Miner Res 2003;18:1261–6.[CrossRef][Medline]
  29. Mughal M, Langton C, Utretch G, Morrison J, Specker B. Comparison between broad-band ultrasound attenuation of the calcaneum and total body bone mineral density in children. Acta Paediatr 1996;85:663–5.[Medline]




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Economos, C D
Right arrow Articles by Naumova, E N
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Economos, C D
Right arrow Articles by Naumova, E N

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
BJR DMFR IMAGING  ALL BIR JOURNALS