Part of this work was presented at the Meeting of the American Society for Bone and Mineral Research, Baltimore, MD, U.S.A., September 9–13, 1995 (Lang T, Fan B, Wu C, et al. 1995 Assessment of vertebral deformation with a high-resolution fanbeam bone densitometer. JBMR 10: M461.)
A Preliminary Evaluation of the Lunar Expert-XL for Bone Densitometry and Vertebral Morphometry†
Article first published online: 1 JAN 1997
Copyright © 1997 ASBMR
Journal of Bone and Mineral Research
Volume 12, Issue 1, pages 136–143, January 1997
How to Cite
Lang, T., Takada, M., Gee, R., Wu, C., Li, J., Hayashi-Clark, C., Schoen, S., March, V. and Genant, H. K. (1997), A Preliminary Evaluation of the Lunar Expert-XL for Bone Densitometry and Vertebral Morphometry. J Bone Miner Res, 12: 136–143. doi: 10.1359/jbmr.19126.96.36.199
- Issue published online: 4 DEC 2009
- Article first published online: 1 JAN 1997
- Manuscript Accepted: 29 AUG 1996
- Manuscript Received: 26 JUL 1996
We have evaluated the Lunar Expert-XL for standard bone mineral densitometry (BMD) and for morphometric imaging of the spine in the lateral projection. The short-term precision in vitro of the Expert-XL for BMD measurements was 0.7% for the Hologic and European spine phantoms, and the long-term stability (15-month) measurements had a 1.1% coefficient of variation. The precision in vivo for three operators examining a group of 10 premenopausal women was 0.9–1.5% for lumbar spine (L2–L4), 1.7–2.2% for the femoral neck, and 0.9% for the total hip region of interest. For a group of nine postmenopausal women, the lumbar spine, femoral neck, and total femur precision ranges were 2.0–2.4, 1.2–2.9, and 1.6–1.7% respectively. For L2–L4, BMD comparison between the Expert-XL and DPX-L yielded a correlation coefficient of r = 0.98, a slope of 0.86, and an intercept of 0.139 g/cm2. The femoral neck results were r = 0.92, slope = 1.00, and intercept = 0.03 g/cm2. In an evaluation of the Expert-XL for lateral morphometry, we employed a group of 16 postmenopausal women. Comparison of vertebral dimensions between the Expert-XL and radiographic morphometry showed strong agreement (r = 0.97), but the interobserver variability for vertebral height was higher for the Expert-XL than for radiographs (3–5% vs. 1–2%). In a subset of four women who had repeat scans, the interscan precision for measuring vertebral dimensions was 1.9, 4.1, and 4.3% for the L1–L4, T12–T8, and T7–T5 levels respectively. In the Expert-XL images, it was possible to identify clearly L4–T4 in 10 of 16 patients and L4–T6 in 15 of 16, indicating potential utility for vertebral fracture prescreening.
Since its inception nearly a decade ago, dual-energy X-ray absorptiometry (DXA) has become a primary noninvasive method for measurement of bone mineral density (BMD) for diagnosis and follow-up of metabolic bone disease.1–6 The wide clinical use of DXA has been based on its low radiation dose, good precision, and the large number of skeletal sites it is capable of measuring. The technique has benefited from continual refinement, reducing both acquisition times and precision errors.7,8 The newest generation of DXA systems is embodied by the Lunar Expert-XL (Lunar Corp., Madison, WI, U.S.A.) and the Hologic QDR-4500 (Hologic, Waltham, MA, U.S.A.). With both of these systems, a high-resolution detector array, coupled to an X-ray tube in a fanbeam geometry, is mounted on a rotateable C-arm gantry. In evaluating the impact of these changes, two major issues must be addressed. On the one hand, densitometric precision errors may be reduced by improved definition of bone edges and intervertebral spaces, while on the other hand, they may be increased by the combination of a fanbeam geometry and the highly complex detector array and its associated electronics. Also, in the lateral projection, these systems may resolve the vertebral endplates more clearly than previous densitometry systems, providing improved evaluation of vertebral fracture.
At our institution, over the last 15 months, we have carried out a performance evaluation of the Lunar Expert-XL bone densitometer. The Expert-XL features a current-integrating detector array with 288 0.8-mm elements mounted on a C-arm opposite a rotating-anode X-ray tube. Dual-energy imaging is achieved by filtration of the X-ray beam and use of detector arrays sensitive to lower and higher energy X-rays. The detector arrays and X-ray tube are mounted on a C-arm gantry that can rotate 90° for supine lateral examinations of the spine.
Because this new system can perform both standard anteroposterior (AP) densitometry and high resolution spine imaging, our performance evaluation contains both densitometric and morphometric studies. To evaluate the implications of this new system for BMD studies, we performed measurements in vitro and in vivo on both Expert-XL and DPX-L systems. Measurements in vitro included short-term and long-term precision as well as BMD response to variations in lean and adipose soft-tissue thickness. In vivo, we evaluated the precision for AP-spine and femur studies as well as the correlation between BMD measurements on the Expert-XL and DPX-L. To evaluate the capability of this system for lateral morphometry, we performed a study in vivo comparing morphometric measurements on the Expert-XL to morphometric measurements from lateral radiographs.
MATERIALS AND METHODS
BMD studies in vitro
Short-term precision of the Expert-XL was measured by 10 repeat scans of a spine phantom (Hologic). Long-term precision was evaluated by daily scans of the same spine phantom (Hologic) over a period of 15 months. Linearity of BMD response for the Expert-XL was measured by 10 repeat scans of the European Spine Phantom (QRM, Mohrendorf, Germany) and calculated as the standard estimate of the error (SEE) of the regression between the measured and the known BMD values of the three vertebral inserts. BMD response to variable soft-tissue thickness was evaluated on both the Expert-XL and DPX-L by imaging different thicknesses of water and acrylic placed uniformly over an aluminum spine phantom (Lunar). BMD response to changes in vertebral fat content was determined on both the Expert-XL and DPX-L by imaging vertebral inserts with 150 mg/cc BMD containing 0, 15, 30, and 50% fat by volume and placed atop a 10-cm layer of acrylic.
For BMD studies, 10 young, healthy women (average age 32 ± 8 years, average BMD 1.18 ± 0.12 g/cm2) and 9 healthy postmenopausal women (average age 63 ± 9 years, average BMD 1.03 ± 0.14 g/cm2) were recruited. For lateral morphometry studies, 16 postmenopausal women (average age 67 ± 14 years) were recruited. For all subjects, informed consent was obtained according to the policies and procedures of the Committee on Human Research of the University of California at San Francisco.
BMD studies in vivo
Subjects underwent repeat supine AP spine and femur imaging, with repositioning between scans, on both the Expert-XL and the DPX-L. On the Expert-XL, all scans were performed in the “fast” mode. On the DPX-L, patients were scanned in the “fast” or “medium” mode based on body thickness according to the guidelines of the manufacturer.
Analysis of BMD image data
The duplicate AP spine and hip scans were analyzed using Expert-XL software version 1.4 and DPX-L version 1.3Z. Analysis of Expert-XL spine images was performed as follows. The bone edges were first carefully evaluated and adjusted for appropriateness in one scan, and subsequently the edges in the second scan were adjusted, if necessary, for consistency with the first scan. Next, the L1–L4 regions of interest (ROIs) in the first scan were placed with careful adjustment of the intervertebral (IV) spaces. The ROIs of the first scan were then mapped to the second scan using the “compare” feature, but the IV spaces were adjusted if necessary. To determine the impact of automated edge detection on the AP-spine precision values, the scans were additionally analyzed with careful adjustment of IV spaces but without adjustment of the bone edges. For Expert-XL AP femur scans, the automated software was employed to determine the bone edges and to place the standard ROIs (femoral neck, trochanter, Ward's triangle, and total hip), but the ischium was manually “painted out” if the femoral neck ROI overlapped with it. To determine the component of precision error due to inter-operator variation of the Expert-XL analysis software, the spine and hip images were analyzed separately and independently by three trained observers.
BMD precision measurements
In vitro, precision was determined as the percentage coefficient of variation (%CV), defined as the standard deviation (SD) of the n measurements normalized by the mean value. In vivo, precision was calculated as the root mean square (RMS) %CV of the pairwise scans as suggested by Glüer.9 Interoperator variation was equivalently calculated as the RMS %CV of the three operators for one set of scans.
Correlation of BMD measurements between the Expert-XL and DPX-L
Interdevice comparison of BMD was performed by linear regression of the means of the paired spine and hip scans on the Expert-XL and DPX-L. The spine comparison was based on the L2–L4 ROI, and the hip comparison was based on the femoral neck ROI.
Acquisition and analysis of lateral spine images
Patients lay supine on the Expert-XL imaging table and had lateral spine scans, which were performed in the “fast mode” (134 kVp, 5 mA, 10 mm/s scan speed) or the “medium mode” (134 kVp, 5 mA, 5 mm/s scan speed) depending on the manufacturer's guidelines for patient thickness. Four of the patients had duplicate Expert-XL scans for assessment of interscan precision. To determine vertebral heights from the Expert-XL images, two experienced observers placed fiducial points at the inferior and superior mid-, anterior, and posterior locations on the vertebral endplates for levels T4–L4. These points were placed interactively on the terminal of the Expert-XL computer workstation. The mid-, anterior, and posterior vertebral heights were then calculated as the distance between the inferior and superior points at each of the three locations. Where either observer determined that the vertebral level was not sufficiently visualized, that level was recorded and excluded from further analysis. Following acquisition of the Expert-XL image, patients had conventional lateral thoracic and lumbar radiographs in the decubitus projection using a technique of 80 kVp, 70 mA, and a 2-s exposure, with a film-to-focus distance of 100 cm. The lumbar and thoracic exposures were centered over L3 and T7, respectively. The radiographs were then placed on a digitizing tablet, and the two observers marked the locations of the six vertebral fiducial points for L4–T4. The posterior mid- and anterior vertebral heights were then computed as before. Additionally, for both Expert-XL and radiographic data sets, the AP, mid-posterior (MP), and mid-anterior (MA) height ratios were calculated. Several comparisons between radiographs and Expert-XL were made.
First, the interobserver variation was computed for each vertebral level both on the radiographs and on the Expert-XL images. For ratios and heights, for each vertebral level, the interobserver variance was computed as the RMS of the differences (RMSD) between the two observers. To compare vertebral heights between modalities, a linear regression was performed between Expert-XL images and radiographs. Comparison of vertebral height ratios was performed by computing the RMSD of height ratios between the two modalities. Interscan precision for the Expert-XL vertebral heights was calculated as the RMS %CV of the duplicate scans on four volunteers. This was calculated separately for the lumbar vertebral levels (16 vertebrae), the lower-thoracic vertebral levels T12–T8 (20 vertebrae), and the upper thoracic levels T7–T5 (9 vertebrae). Finally, it should be noted that although the two observers prospectively agreed on general guidelines for evaluation of the images, specific cases were analyzed separately without discussion between them.
BMD studies in vitro
Short-term precision of the Expert-XL was 0.7% for both the Hologic and European Spine Phantoms. The long-term (15-month) in vitro BMD measurements had a 1.1% overall CV (Fig. 1), which was larger than normally observed for Lunar DPX systems as monitored in clinical BMD trials by our laboratory. The variance could be attributed to significant discontinuities, or shifts of the mean, in the BMD measurements as identified by a cumulative sum (CUSUM) analysis.10 Scans of the European spine phantom showed that the response of the Expert-XL to changes in BMD was highly linear, giving a 0.006 g/cm2 SEE. Response of Expert-XL BMD measurements to differences in thicknesses of fat, water, and acrylic are demonstrated in Figs. 2, 3, 4. The fat dependence of the two systems were almost identical and were consistent with the 0.012 g/cm2 value reported by Glüer for the QDR-1000 under identical conditions.11 There was a statistically significant dependence of BMD on the thickness of acrylic but not on the thickness of water, and this held for both the DPX-L and the Expert-XL. There was no significant dependence of BMD on variation of table height.
BMD studies in vivo
Precision values obtained by each observer for Expert-XL spine and hip ROIs are summarized in Table 1. Precision values for the corresponding measurements on the DPX-L are summarized in Table 2. When the bone edges calculated by the Expert-XL spine software were not modified, the L1–L4 precision values obtained by observers 1 and 2 worsened from 0.9–1.1% to 1.5–1.8%. Interobserver variations for the Expert-XL spine and hip ROIs are summarized in Table 3. Figures 5a and 5b represent the degree of intersystem correlation observed for the L2–L4 and femoral neck ROIs.
Lateral spine imaging
Figure 6 is an image of a 51-year-old volunteer in which vertebral levels from L4 to T4 could be visualized. Generally, the Expert-XL images showed clear definition of the lumbar and lower thoracic vertebral endplates, but the upper thoracic vertebral endplates could not be defined in some of the patients due to low contrast and the presence of artifacts from overlying tissues. The effect of variable thoracic image quality is described in Table 4, which summarizes the distribution of highest analyzable vertebral levels for the two observers. It should be noted that the two observers' results agreed on identification of excluded levels in 10 of 16 images. Interobserver variations for vertebral heights are compared for Expert-XL images and conventional radiographs in Tables 5 and 6. The corresponding values for vertebral height ratios were very similar. The interobserver variation for the Expert-XL was generally higher than that for radiographs, but neither technique showed a clear dependence on vertebral level. Regression analysis showed a highly linear correlation between vertebral heights measured on the two modalities. Although Fig. 7 demonstrates this relationship for anterior heights, the same relation held true for posterior and mid-vertebral dimensions. Table 7 shows the RMSD for vertebral height ratios measured on the Expert-XL and conventional radiographs. Again, the results were independent of vertebral level. Finally, the interscan precision for vertebral heights was 1.9, 4.1, and 4.3% for the lumbar vertebrae, lower-thoracic, and upper-thoracic vertebrae, respectively.
Although the use of a fanbeam geometry for BMD and lateral spine measurements was introduced initially on the QDR-2000 (Hologic),7,12 the design of the Expert-XL, particularly as regards the detector array, X-ray tube, and gantry, is sufficiently novel to merit a thorough review of implications of these changes for BMD and lateral morphometry measurements. The spatial resolution and noise properties of the Expert-XL densitometric images are unique to this system and have significant influence on the precision of the BMD measurements. Thus, this report was focused on outlining key sources of precision error, such as edge detection and interoperator variation. To understand the implications of replacing older systems with the Expert-XL, it is also important to compare Expert-XL BMD measurements with those of current technology. Thus, this work compared in vivo AP spine and hip measurements between the Expert-XL and the DPX-L, as well as in vitro response to variations in fat and soft-tissue thickness. In addition to BMD measurements, the detector, X-ray tube, and gantry design of the Expert-XL permit supine examinations of the spine in the lateral projection on a 30-s timescale. The combination of short scan times and improved ability to delineate the vertebral endplates implies that the Expert-XL has significant potential to perform vertebral fracture assessment and morphometry in a clinical setting. Thus, to understand the implications of using this new technology, it is important to assess how morphometric measurements made by the Expert-XL compare with those made from standard radiographs and to assess comparatively the interoperator precision of the two techniques.
Our BMD precision measurements showed that the performance of the Expert-XL for AP hip scanning was comparable to or better than the DPX-L, but that the spine precision errors were somewhat larger than the DPX-L. For the whole study group, the Expert-XL had a L2–L4 precision of 1.7–1.9% compared with 1.3% for the DPX-L. This difference may originate from several factors. One major factor may be the need for further adjustment of the spine edge-detection algorithm. In evaluation of the spine data, we found that it was necessary to adjust the bone edges for accuracy in most (approximately two-thirds) of the images. Because of the operator intervention thus required, the BMD precision results varied between observers and ranged between 0.9 and 1.4% for young volunteers and 2–2.5% for older women, with the result probably depending on the degree of care in edge adjustment taken by each operator. That edge detection may be a key aspect of the spine precision error is supported by the interoperator precision value of 0.6%. Another source of variability in the spine data may be the composition of the regions employed by the Expert-XL software as a soft-tissue baseline. The short (12-s) scan time may imply that these regions are more sensitive to effects such as bowel gas or breathing artifacts, which may be smoothed over in longer acquisitions. Short-term variations in the scanner BMD measurements may also contribute, given our short-term in vitro precision measurements of 0.7% on the Hologic and ESP spine phantoms.
In the hip, where precision errors are dominated by patient positioning and by variable placement of the neck ROI, rather than by edge detection or variable soft-tissue composition, the Expert-XL was comparable to or better than the DPX-L. Of particular interest is the total hip measurement, which showed a precision better than 1% in young volunteers. This high precision value relates to the size of the ROI and the ease of edge detection in the region. Finally, the long-term precision measurements, which showed significant variability in the measurement of BMD over time, represent an additional area of concern. The large overall CV (1.1%) may be partly attributed to the fact that the Expert-XL in our laboratory was the first to be installed outside of the factory and was subject to several repairs and upgrades during its first year of operation. However, there were few outliers or resolvable short-term oscillations in these measurements, indicating that the calibration system was successfully correcting short-term errors, but that the correction mechanism itself was drifting over time.
There was a reasonable correspondence between the Expert-XL and the DPX-L as evidenced by in vitro studies as well as by comparison of in vivo spine and hip BMD measurements. The Expert-XL and DPX-L had similar responses to variable fat- and soft-tissue thickness, and the results for both systems point to the limitations in the use of BMD measurements in obese subjects13–15 or in subjects who have experienced a large weight gain between visits in a longitudinal study. For AP spine studies, although the measurements were highly correlated (r = 0.98, CV = 2.5%), there was a significant slope and intercept (slope = 0.86, intercept = 0.14 g/cm), indicating a possible difference in calibration between the two machines. The hip results showed more scatter (r = 0.92, CV = 5.2%) but a nonsignificant slope and intercept. The large CV may have been due to the effect of variable positioning of the neck box between the two systems. It should be noted that the comparisons employed mean values of the paired scans rather than single measurements, thus increasing the correlation coefficients by reducing the effect of precision errors. The correlation results thus obtained are consistent with the generally accepted practice of not changing densitometers during longitudinal evaluations of single patients or of groups of patients in clinical trials.
The potential utility of the Expert-XL for lateral spine imaging is based on the ability to use the same instrument both to perform standard densitometric examinations and to diagnose vertebral fractures at a lower radiation dose than conventional spinal radiographs. Two questions must be comparatively assessed. The first issue is the ability of the two techniques to diagnose qualitatively vertebral fractures. In our study, both observers could clearly visualize vertebral levels from L4–T4 from the spinal radiographs in all of the patients. In the Expert-XL images, it was possible to visualize this range of levels in 10 of 16 patients and from L4–T6 in 15 of 16 patients. The utility of the Expert-XL for qualitative diagnosis of vertebral fracture was also addressed by Hans,16 who found reasonable agreement, as evidenced by a κ-score of 0.86, between consensus readings of radiographs and qualitative assessment of fractures from Expert-XL images in a group of 21 osteoporotic patients. The second issue that was addressed in this study is comparison between quantitative vertebral measurements with the two techniques. Our study, which assessed the intertechnique differences between quantitative morphometric measurements in a set of relatively healthy volunteers, showed that the differences between techniques, expressed as the RMSD of vertebral height ratios and the SEE of regressions comparing vertebral height measurements, were on the order of 2–5%. We found that interobserver and interscan precision for the Expert-XL had a similar magnitude, but also that the interobserver errors for conventional radiographic morphometry were somewhat lower, ranging from 1–3%. It should be noted that the intertechnique and interobserver variations for the Expert-XL are generally consistent with precision values reported in the literature for radiographic morphometry17 and are small compared with the 20% reduction in height and/or height ratios, which is often used as criteria for defining moderate vertebral fractures.
However, it should be noted that the Expert-XL interoperator and precision errors documented here are probably lower than would be the case in routine clinical usage. This study was conducted in a group of healthy postmenopausal women, whereas in an osteoporotic population, low BMD would result in both higher precision errors and weaker correlations with conventional radiography due to reduced ability to visualize the vertebral endplates. Another factor that tended to reduce precision errors and increase the correlation with conventional morphometry was that vertebral levels not identifiable by both observers on the Expert-XL images were excluded from this analysis.
In conclusion, the Expert-XL has advantages and limitations as a clinical imaging system. The advantages are the short examination times, which promote patient comfort and economy of usage, and the high image resolution, which permits detailed examination of the lateral spine. The densitometric precision errors are small compared with the population variations in BMD, meaning that the system can be effectively used for fracture risk screening. However, in longitudinal BMD examinations of the lumbar spine, it is important to monitor carefully the bone edges to ensure acceptable precision. Thus, lumbar spine BMD measurements are an aspect of the system that will require further refinement by the manufacturer.
The use of the Expert-XL for lateral morphometry represents an interesting new dimension for densitometry in that it offers the possibility to diagnose vertebral fracture at a substantially reduced radiation dose (40 μSv18 vs. 700 μSv19) compared with lateral lumbar radiographs. Our preliminary measurements in a group of postmenopausal women indicate that it may be possible to employ the Expert-XL to screen for vertebral fracture in levels from L4 to T4 in most of the population. However, in a significant number of cases, an additional radiographic measurement would be required to establish the fracture status in the upper thoracic levels not visualized in the Expert-XL scan. Even so, use of the system would result in a substantial reduction of radiation dose for those cases in which the L4–T4 endplates were successfully visualized and a small incremental dose for those cases where an additional radiograph was necessary. While the Expert-XL system thus has obvious promise for qualitative assessment of vertebral fracture, further research is required to establish the utility of the system for more quantitative measurements. Due to reduced spatial resolution and higher image noise, the interobserver and interscan variabilities are higher for the Expert-XL than for radiography, and it is important to understand the impact of these errors both on the detailed classification of vertebral fracture severity as well as on serial measurements of vertebral deformity. Advances in these capabilities may result from improvements in image processing and in scanning protocols, particularly as they apply to the thoracic component of the Expert-XL scan.
This research was partially supported by the Lunar Corporation. The authors thank Dr. Thomas Fuerst, Dr. Claus Glüer, and Dr. Ying Lu for technical input. The authors are grateful to Dr. James Hanson, Mr. Joseph Bisek, and Mr. Chris Fowler for technical support.
- 11992 Assessment of spinal and femoral done density by dual x-ray absorptiometry: Comparison of Lunar and Hologic instruments J Bone Miner Res 7: 1081–1084., , , , ,
- 21994 Dual-energy x-ray absorptiometry of the spine in anteroposterior and lateral projections Osteoporosis Int 4: 110–116., , ,
- 31993 Bone geometry, structure, and mineral distribution using DXA Calcif Tissue Int 52: 154.,
- 41995 Recent advances in the non-invasive assessment of bone density, quality and structure J Jpn Soc Bone Morphom 5: 93–108., , , ,
- 51993 Current methods and recent advances in the diagnosis of osteoporosis Arthrit Rheumat 36: 1649–1662.,
- 61995 Dual-energy x-ray absorptiometry for the management of bone disease Phys Med Rehab Clin North Am 6: 507–537.
- 71991 Paired AP and lateral supine dual x-ray absorptiometry of the spine: Initial results with a 32 detector system Osteoporosis Int 1: 190., , ,
- 81992 Enhanced precision with dual-energy x-ray absorptiometry Calcif Tissue Int 51: 14–17., , ,
- 91995 Accurate assessment of precision errors: How to measure the reproducibility of bone densitometry techniques Osteoporosis Int 5: 262–270., , , ,
- 101996 Dual-energy X-ray absorptiometry quality control: Comparison of visual examination and process control charts J Bone Miner Res 11: 626–637., , , , , , , et al.
- 111990 Comparative assessment of dual-photon-absorptiometry and dual-energy-radiography Radiology 174: 223–228., , , , ,
- 121994 Morphometric X-ray absorptiometry of the spine: Correlation in vivo with morphometric radiography Osteoporosis Int 4: 238–244., , ,
- 131988 Comparison of dual-energy x-ray absorptiometry and dual photon absorptiometry for bone mineral measurements of the lumbar spine Mayo Clin Proc 63: 1075–1084., , , ,
- 141991 Errors in DXA of spine due to non-uniform fat distribution Osteoporosis Int 1: 199.,
- 151991 Precision and stability of dual energy X-ray absortiometry measurements Calcif Tissue Int 49: 174–178.,
- 161995 Expert-XL: Clinical evaluation of a new morphometric technique on 21 patients with verteral fracture. 11th International Bone Densitometry Workshop, Gleneden Beach, OR, U.S.A., , , , ,
- 171995 Comparison of quantitative and semi-quantitative techniques for the assessment of vertebral fractures Osteoporosis International 5: 354–370., , ,
- 18(Lunar 1996 Expert-XL Users Manual for software version v1.52. Lunar Corp., Madison, WI, U.S.A.
- 191992 Effective dose values in bone mineral measurements by photon absorptiometry and computed tomography Osteoporosis Int 2: 82–87.