Abstract
Objectives
To evaluate the influence of intravenous contrast agent on the diagnostic ability for osteoporosis using CT attenuation measurement in patients with liver cirrhosis.
Methods
This retrospective study was approved by our institutional review board and informed consent was waived. 208 patients with liver cirrhosis (mean age, 61.25 years ± 9.43 [standard deviation]; range, 30–82 years) who underwent both unenhanced and two contrast-enhanced (arterial and venous phase) abdominal dual-energy CT examinations from January 1 to September 1, 2020, were recruited. CT attenuation values were measured in the medullary compartment of vertebral body (L1–L3) and bone mass was determined by the hydroxyapatite concentration obtained in dual-energy spectral CT and used as the reference standard. Receiver operating characteristic (ROC) curves were used to evaluate the diagnostic ability of using CT attenuation number in unenhanced, arterial, and venous phases.
Results
Area under ROC curve using unenhanced CT attenuation was different from using arterial CT attenuation (p
= 0.038) and venous CT attenuation (p = 0.048) to diagnosing osteoporosis. However, there was no significant difference between unenhanced CT attenuation and arterial CT attenuation (p = 0.773) and between unenhanced CT attenuation and venous CT attenuation (p = 0.746) to distinguish low bone mass (osteoporosis or osteopenia).
Conclusions
The diagnostic ability for osteoporosis using CT attenuation measurement in contrast-enhanced scans is decreased due to intravenous contrast contamination, however, which had no influence on the diagnostic ability of CT attenuation for low bone mass (osteoporosis or osteopenia).
Advances in knowledge
The diagnostic ability of using enhanced CT attenuation values for osteoporosis decreased compared to unenhanced CT attenuation values.
Hepatic cirrhosis is becoming a more common medical problem. Metabolic abnormalities result in various comorbidities, including a higher risk of osteoporosis. 1 Chest and abdominal CT can be used to screen osteoporosis. 2–4 However, many body CT examinations are solely performed with intravenous contrast medium in clinical practice. Some studies reported that the diagnostic ability to predict osteoporosis was similar between enhanced and unenhanced examinations. 5,6 However, another study showed contrast agent on enhanced CT scan may underestimate osteoporosis compared to unenhanced CT examinations. 7 It is thus not certain whether the bone mass estimation would be affected by contrast agent and whether using enhanced CT attenuation values have the same diagnostic ability for screening osteoporosis to unenhanced CT attenuation values.
Dual-energy spectral CT has been reported to have the ability to measure material concentration, including hydroxyapatite concentration. 8–11 Furthermore, dual-energy spectral CT could accurately assess hydroxyapatite with smaller bias in bone mineral density evaluation, compared to conventional quantitative CT. 12
Thus, the aim of this research was to evaluate the influence of intravenous contrast agent on the accuracy of estimating bone mass quantified by hydroxyapatite (HAP) concentration in dual-energy spectral CT in cancellous bone of lumbar vertebral bodies (L1–L3) in patients with liver cirrhosis using CT attenuation measurements in different scan phases and thus the influence of intravenous contrast agent on the diagnostic ability for osteoporosis.
Methods and materials
Study population
Our study was performed with the approval of our institutional review board. Informed consent was waived for the retrospective analysis. The radiological information and picture archiving and communication system (PACS) from January 1 to September 1, 2020, was queried for patients eligible for study inclusion fulfilling following criteria 1 : Patients with liver cirrhosis. 2 Patients underwent both the unenhanced and the two contrast-enhanced (arterial and venous phase) abdominal dual-energy spectral CT phases. Exclusion criteria: Patients were excluded if vertebrae L1–L3 were involved with metastasis.
All patients had clinically confirmed liver cirrhosis with portal hypertension by clinical history, laboratory, ultrasound, or MRI, and were definitively diagnosed with liver cirrhosis according to the Viral Hepatitis Prevention and Control Programs. 13 The severity of liver cirrhosis was evaluated with the Child–Pugh classification. There were 70 patients in Child–Pugh Grade A, 90 in Grade B, and 48 in Grade C. All patients underwent dual-energy spectral CT for evaluating the hepatic abnormality and the status of portal-systemic circulation. Eleven patients who were diagnosed as or suspicious for bone metastasis were excluded. Bone metastasis was defined as lesion that had indeterminate morphologic features with ill-defined margins, increased radionuclide uptake on bone scintigraphs (or PET-CT), and an increase in the size and/or density on follow-up CT images (one patient in our study). Any equivocal bone metastasis lesion was defined as 1 positive on bone scintigraphs (or PET-CT) but no change in the size and density on follow-up CT images (nine patients in our study) or 2 negative on bone scintigraphs (or PET-CT) but an increase in the size and/or density on follow-up CT images (one patient in our study). 14
Ultimately, among 219 patients potentially eligible for study inclusion, 208 patients were included (male/female, 105/103).
Risk factors for bone disease were determined, which included age, weight, height, body mass index, and so on, as depicted in Table 1.
Table 1.
Clinical characteristics of patients classified by hydroxyapatite concentration measured on spectral CT
| Characteristic | Patients with normal bone mass | Patients with osteopenia | Patients with osteoporosis | P value |
|---|---|---|---|---|
| n | 36 | 86 | 86 | |
| Male/female* | 18/18 | 41/45 | 39/47 | 0.887 |
| Age (years) | 51.83 ± 9.28 | 60.15 ± 7.96 | 66.29 ± 7.28 | <0.05 |
| Body weight (kg) | 61.21 ± 6.13 | 62.88 ± 7.95 | 60.61 ± 6.97 | <0.05 |
| BMI (kg/m2) | 22.76 ± 2.13 | 23.05 ± 2.22 | 22.39 ± 2.06 | <0.05 |
| Etiology of the cirrhosis* | 0.998 | |||
| Viral-related | 22 | 53 | 55 | |
| Alcoholic-related | 8 | 18 | 18 | |
| Biliary cirrhosis | 2 | 6 | 6 | |
| Other | 4 | 9 | 7 | |
| Child–Pugh Grade | 0.247 | |||
| Child–Pugh A | 18 | 35 | 17 | |
| Child–Pugh B | 28 | 32 | 30 | |
| Child–Pugh C | 15 | 15 | 18 |
BMI, body mass index.
Note—Unless otherwise indicated, data are means ± standard deviation. Adopting standard on quantitative CT for abnormal bone mass of vertebral trabecular bone, thresholds of 80 mg/cm3 of hydroxyapatite for osteoporosis and 120 mg/cm3 of hydroxyapatite for osteopenia were also used in our study.
CT examinations
All patients underwent routine abdominal dual-energy spectral CT with a 256-row multislice CT system (third-generation dual-energy CT system, GE Revolution GSI, GE Healthcare, Chicago, Illinois, USA). Scanning covered abdomen ranged from diaphragmatic dome to both kidneys. The scanning parameters included: 0.984 of helical pitch; 0.8 s of gantry rotation time; 0.625 mm of axial section thickness; 40 mm of collimator width; 30 cm of display field of view. Contrast agent (Bayer Vital, Berlin, Germany, IC: 450 mg iodine/ml) was administered via the antecubital vein using a standard dosage of 1.2 ml/kg of body weight at a flow rate of 4 ml s−1, followed by 30 ml of saline solution at the same flow rate for 20 s. The bolus-tracking technique was used to trigger the arterial phase scan. The region of interest (ROI) for monitoring the contrast enhancement was placed in the descending aorta, and the arterial phase scan started 7 s after the CT attenuation value in the ROI reached 100 HU. The portal venous phase scan started 30 s after the finish of arterial phase scan.
Data processing
Both the HAP-based material decomposition images and monochromatic images were reconstructed. These CT images were analyzed with an advanced workstation (AW4.7; GE Healthcare) using a software package (AW Volume Share seven and AW4.7 Ext. 16 Software; Gemstone Spectral Imaging Viewer; GE Healthcare). The HAP concentrations were measured on the HAP-based material decomposition images and attenuation value on the 70keV monochromatic CT images. Adopting standards on quantitative CT for abnormal bone mass of vertebral trabecular bone, thresholds of 80 mg/cm3 of HAP concentration for osteoporosis and 120 mg/cm3 of HAP concentration for osteopenia were also used in our study. 15 Only HAP concentrations at the unenhanced phase were used as the reference standards of bone mass in our study.
All regions of interest placement on lumbar vertebral bodies (L1–L3) were performed by one reader (Reader 1, an orthopedic radiologist with 15 years of experience). The field-of-view for obtaining the measurements in the vertebral bodies was standardized; so every image was magnified with display field-of-view 17.0 cm * 17.0 cm for precise measurement in our workstation. The Basivertebral vein entry was used to confirm the section for analysis, as previously described. 8 The area and location of region of interest in the two enhanced phases were the same as the unenhanced phase (Figure 1). Region of interest was placed three times in each patient, and the values from the three measurements were averaged to represent the final values. If a patient suffered a fracture or deformity on one vertebral body, the measurement was made on the adjacent vertebral body.
Figure 1.
Spectral CT examination of 30 year old male with hepatitis B virus-related liver cirrhosis. (a) Lumbar spine in sagittal CT reformation. Note compression fractures in T11, T12, L1, and L2. Basivertebral vein entry is used to confirm level for analysis. (b) Hydroxyapatite concentration was 123.3 mg/cm3 for L3. (c) CT attenuation value was 200.5 Hounsfield in the unenhanced phase for L3. (d) CT attenuation value was 216.8 Hounsfield in the arterial phase for L3. (e) CT attenuation value was 236.9 Hounsfield in the venous phase for L3
Variation
To test variation, we chose 20 patients in a random manner. The CT attenuation values of the lumbar vertebral bodies were measured by two readers (Reader 1, and Reader 2, an orthopedic radiologist with 30 years of experience) independently. Variation for CT attenuation measurements between the two results and those of the regular measurement by Reader 1 were analyzed.
Statistical analysis
Continuous variables were expressed as means ± standard deviation and categorical values were expressed by percentages. We used a post hoc analysis to compare CT attenuation values of the unenhanced, arterial phase, and venous phase. The receiver operating characteristic (ROC) curves were used to evaluate the diagnostic ability of using unenhanced, arterial, and venous phase CT attenuation for estimating bone mass quantified by HAP concentration in dual-energy spectral CT. Tests for differences between pairs of ROC curves were executed using the method described by Hanley and McNeil. 16 A P-value of less than 0.05 was considered to indicate a statistically significant difference. To test interobserver and intraobserver agreements, we performed Pearson correlation coefficients. All statistical tests were performed with the statistical package for social sciences for Windows (SPSS, v.17.0; SPSS, Chicago, Ill).
Results
Participant characteristic
Based on the HAP concentration measured by dual-energy spectral CT, 86 patients were diagnosed with osteoporosis (with HAP concentration below 80 mg/cm3), 86 patients were diagnosed with osteopenia (with HAP concentration from 80to 120 mg/cm3), and 36 patients had normal bone mass (with HAP concentration above 120 mg/cm3). Clinical data of 208 patients are shown in Table 1. For patients with osteoporosis, osteopenia, and normal bone mass, vertebral bone CT attenuation values increased in the venous phase, compared to unenhanced phase and arterial phase. Post hoc analysis showed that significant differences were presented among all three phases in three subgroups (p < 0.001) (Table 2).
Table 2.
Changes in vertebral CT attenuation values between unenhanced and enhanced examinations of patients classified by hydroxyapatite concentration measured on spectral CT
| Unenhanced phase | Arterial phase | Venous phase | |
|---|---|---|---|
| Patients with normal bone mass | 205.29 ± 42.29 | 223.90 ± 43.09 | 234.17 ± 44.90 |
| Patients involved with osteopenia | 139.72 ± 22.90 | 156.08 ± 26.04 | 164.06 ± 26.39 |
| Patients involved with osteoporosis | 87.86 ± 24.85 | 102.47 ± 28.28 | 110.77 ± 28.81 |
Mean ± standard deviation. Post hoc analysis showed that significant differences were presented among all three phases in three subgroups (p < 0.001).
In our study, the use of unenhanced CT attenuation could predict osteoporosis with high area under ROC curve (AUC = 0.957) and the threshold of 119.5 HU could induce high sensitivity (91.86 %), high specificity (87.7 %), high positive-predictive value (84.0 %), and high negative-predictive value (93.9 %). The use of CT attenuation values at the arterial phase could predict osteoporosis with high area under ROC curve (AUC = 0.944) and the threshold of 136.2 HU could induce high sensitivity (90.70 %), high specificity (85.25 %), high positive-predictive value (81.2 %), and high negative-predictive value (92.9 %). The use of CT attenuation value in the venous phase could predict osteoporosis with AUC of 0.944 in ROC and the threshold of 145.1 HU could induce high sensitivity (90.70 %), high specificity (84.43 %), high positive-predictive value (80.4 %), and high negative-predictive value (92.8 %). The AUC value of using unenhanced CT attenuation was different from that of using arterial CT attenuation to diagnose osteoporosis (p = 0.038) (Figure 2a). Likewise, AUC of unenhanced CT attenuation was different from venous CT attenuation to diagnose osteoporosis (p = 0.048) (Figure 2b). However, there was no significant difference between the use of arterial CT attenuation and venous CT attenuation in diagnosing osteoporosis (p = 0.939).
Figure 2.
Receiver operating characteristic (ROC) curves are used to distinguish osteoporosis in patients with liver cirrhosis.(a) Unenhanced CT attenuation was different from arterial phase CT attenuation to distinguish osteoporosis (p = 0.038). (b) Unenhanced CT attenuation was different from venous phase CT attenuation to distinguish osteoporosis (p = 0.048)
Furthermore, unenhanced CT attenuation could predict low bone mass (osteoporosis or osteopenia) with a high AUC value of 0.964 and the threshold of 155.4 HU provided high sensitivity (87.21 %), high specificity (97.22 %), high positive-predictive value (99.3 %), and moderate negative-predictive value (61.4 %). While the arterial CT attenuation value had AUC of 0.962 and with the threshold of 169.6 HU provided high sensitivity (84.88 %), high specificity (94.44 %), high positive-predictive value (98.6 %), but low negative-predictive value (56.7 %). Venous CT attenuation also had high AUC value of 0.961 and with the threshold of 189.4 HU provided high sensitivity (93.02 %), high specificity (88.89 %), high positive-predictive value (97.6 %), and good negative-predictive value (72.7 %). There was no significant difference between unenhanced CT attenuation and arterial CT attenuation (p = 0.773) (Figure 3a) and between unenhanced CT attenuation and venous CT attenuation to distinguish low bone mass (osteoporosis or osteopenia) (p = 0.746) (Figure 3b). In addition, there was no significant difference between the arterial CT attenuation and venous CT attenuation in differentiating low bone mass (osteoporosis or osteopenia (p = 0.940).
Figure 3.
Receiver operating characteristic (ROC) curves are used to distinguish low bone mass (osteoporosis or osteopenia) in patients with liver cirrhosis.(a) Unenhanced CT attenuation was not significantly different from arterial phase CT attenuation to distinguish low bone mass (osteoporosis or osteopenia) (p = 0.773). (b) Unenhanced CT attenuation was not significantly different from venous phase CT attenuation to distinguish low bone mass (osteoporosis or osteopenia) (p = 0.746)
Variation
There were excellent interobserver and intraobserver agreements for bone attenuation value measurements in the unenhanced phase, arterial phase, and venous phase (all r > 0.90, p < 0.001).
Discussion
In our study, we investigated the influence of intravenous contrast agent on the accuracy of estimating bone mass and on the diagnostic ability for osteoporosis. Our study showed that contrast enhancement elevated the attenuation in the vertebra, requiring the use of higher CT attenuation thresholds for differentiation diagnosis in the contrast-enhancement phases, and that the diagnostic ability of using CT attenuation values for osteoporosis in the contrast-enhanced phases was decreased in patients with liver cirrhosis. However, the diagnostic ability of CT attenuation for low bone mass (osteoporosis or osteopenia) was not influenced by intravenous contrast agent.
Osteoporosis is a common complication of chronic liver disease, especially in the final stages. It was reported that osteoporosis occurs in 12–55% of patients with liver cirrhosis. 17 Some previous studies investigated the use of CT attenuation values in the unenhanced and contrast-enhanced scan phases in diagnosing osteoporosis and fractures. 18–20 The results were inconsistent in terms of inconclusive whether enhanced CT attenuation values had the same diagnostic ability for screening osteoporosis compared to unenhanced CT attenuation values. Baum et al found that attenuation values had potential to assess bone mineral density. 19 Papadakis et al considered that attenuation values could be used to be a diagnostic tool for osteoporosis. 20 However, the two above studies only used contrast-enhanced CT and did not compare these values with unenhanced values. 19,20 In our study, we found that there was no significant difference between unenhanced CT attenuation and arterial or venous CT attenuation to distinguish low bone mass (osteoporosis or osteopenia), when the attenuation value threshold was adjusted based on the scan phase. However, AUC in ROC curve of using the unenhanced CT attenuation was statistically higher than using arterial CT attenuation or venous CT attenuation to distinguish osteoporosis. Since CT attenuation values in the vertebra in contrast-enhanced phases included not only information of bone mineral density but also information of contrast agent in bone tissue. As we know, tissue enhancement degree was complicated and relied on several indexes, such as cardiovascular disease, tissue blood volume, tissue perfusion, and so on. 21 The contrast enhancement contamination not only required the use of increased thresholds for the differentiation task but also broadened the attenuation value distribution in patients and increased the overlap of attenuation values among different patient groups. The broadened attenuation distribution may be used to explain why diagnostic ability for osteoporosis of the two contrast-enhanced CT attenuation values was lower than that of unenhanced CT attenuation values in our study.
A previous study showed that enhanced abdominopelvic CT was essentially equivalent to unenhanced CT on bone mineral density assessment of femoral neck. 22 The result was different from our study. The reason may be due to different race, patient group, estimation location, and scan protocols. Enhanced attenuation values strongly depend on the scan time after contrast agent injection. In our study, the arterial phase images were acquired at 7 s after the descending aortic enhancement reached 100 HU and the portal venous phase images were acquired at 30 s after the arterial phase scan. In the previous study, 22 the enhanced scan protocols included a split-bolus contrast-injection technique with an injection of 50 ml of contrast material followed by a 10-min delay and subsequent second injection of 100 ml of contrast material with scanning occurring 100 s after the second bolus. Different from our study, David et al used the spectral CT-derived virtual non-contrast images and showed that the use of virtual non-contrast images systematically underestimated volumetric bone mineral density and therefore should not be used without appropriate adjustments. Adjusted venous phase images could provide reliable results and may be utilized for an opportunistic bone mineral density screening in CT examinations. 23 Attenuation in HU depends on a number of factors and different measurement methods may be the reason why there was different conclusions between David et al and ours. Furthermore, all above facts implied that dual-energy radiographic absorptiometry or quantitative CT still require calibration when CT attenuation is used to screen bone mass.
Our feasibility study had some limitations. Firstly, our study was not a randomized-controlled study. Secondly, tissue enhancement degree depends on several indexes, such as cardiovascular disease, tissue blood volume, tissue perfusion, and so on, 21 although we used the method of bolus tracking to alleviate the effect of cardiovascular function on tissue blood supply, external variations existed to influence attenuation measurement in our study. Thirdly, our results in patient involved with liver cirrhosis need to be validated in patient groups of other indications. Fourthly, a single reader performed region of interest placement; however, variability here was measured using a small study of 20 cases using a Pearson correlation coefficient. Fifthly, using attenuation in HU is a proxy depending on a number of factors, so even if statistically there was some correlation of change in attenuation, for a single patient the measurement of bone mineral density would be far more useful. Sixthly, our study referred to hydroxyapatite concentration as a surrogate and a marker of bone mineral density without comparison with dual-energy radiographic absorptiometry. It was one of our limitations, although Mei et al found that high correlations were found between cancellous bone mineral density values assessed with spectral CT and conventional quantitative CT. 24 Lastly, although spectral CT could give relatively accurate material concentration, it was not a gold standard as quantitative CT.
In conclusion, contrast injection elevates the attenuation in the vertebra and decreases the diagnostic ability of using CT attenuation for osteoporosis. However, the diagnostic ability of using CT attenuation for low bone mass (osteoporosis or osteopenia) is not influenced by intravenous contrast agent.
Footnotes
Acknowledgements: Taking this opportunity, the authors express sincere appreciation to the people who gave our team the great support. Yang Zhang, MA, carried out proof reading and improved the language of this paper. This paper cannot conclude without their help and contribution. Hereby again, the study team addresses our sincere gratitude.
Contributor Information
Xinmeng Hou, Email: caomingliuyan@163.com.
Xiaoyue Cheng, Email: cxiaotu@163.com.
Yuangang You, Email: yygg345@163.com.
Jianying Li, Email: Jianying.Li@med.ge.com.
Daqing Ma, Email: madaqing@263.net.
Zhenghan Yang, Email: yangzhenghan@vip.163.com.
Qiang Ma, Email: spenserma@163.com.
REFERENCES
- 1. Wakolbinger R, Muschitz C, Scheriau G, Bodlaj G, Kocijan R, Feichtinger X, et al. Bone microarchitecture and bone turnover in hepatic cirrhosis. Osteoporos Int 2019; 30: 1195–1204. doi: 10.1007/s00198-019-04870-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Lee SJ, Graffy PM, Zea RD, Ziemlewicz TJ, Pickhardt PJ. Future osteoporotic fracture risk related to lumbar vertebral trabecular attenuation measured at routine body CT. J Bone Miner Res 2018; 33: 860–67. doi: 10.1002/jbmr.3383 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Johannesdottir F, Allaire B, Bouxsein ML. Fracture prediction by computed tomography and finite element analysis: current and future perspectives. Curr Osteoporos Rep 2018; 16: 411–22. doi: 10.1007/s11914-018-0450-z [DOI] [PubMed] [Google Scholar]
- 4. Link TM, Koppers BB, Licht T, Bauer J, Lu Y, Rummeny EJ. In vitro and in vivo spiral CT to determine bone mineral density: initial experience in patients at risk for osteoporosis. Radiology 2004; 231: 805–11. doi: 10.1148/radiol.2313030325 [DOI] [PubMed] [Google Scholar]
- 5. Pickhardt PJ, Lauder T, Pooler BD, Muñoz Del Rio A, Rosas H, Bruce RJ, et al. Effect of IV contrast on lumbar trabecular attenuation at routine abdominal CT: correlation with DXA and implications for opportunistic osteoporosis screening. Osteoporos Int 2016; 27: 147–52. doi: 10.1007/s00198-015-3224-9 [DOI] [PubMed] [Google Scholar]
- 6. Jang S, Graffy PM, Ziemlewicz TJ, Lee SJ, Summers RM, Pickhardt PJ. Opportunistic osteoporosis screening at routine abdominal and thoracic CT: normative L1 trabecular attenuation values in more than 20 000 adults. Radiology 2019; 291: 360–67. doi: 10.1148/radiol.2019181648 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Pompe E, Willemink MJ, Dijkhuis GR, Verhaar HJJ, Mohamed Hoesein FAA, de Jong PA. Intravenous contrast injection significantly affects bone mineral density measured on CT. Eur Radiol 2015; 25: 283–89. doi: 10.1007/s00330-014-3408-2 [DOI] [PubMed] [Google Scholar]
- 8. Ma Q, Yang Z, Han X, Liu F, Su D, Xing H. Influence of parathyroidectomy on bone calcium concentration: evaluation with spectral CT in patients with secondary hyperparathyroidism undergoing hemodialysis-A prospective feasibility study. Radiology 2017; 284: 143–52. doi: 10.1148/radiol.2016161797 [DOI] [PubMed] [Google Scholar]
- 9. Graser A, Johnson TRC, Chandarana H, Macari M. Dual energy CT: preliminary observations and potential clinical applications in the abdomen. Eur Radiol 2009; 19: 13–23. doi: 10.1007/s00330-008-1122-7 [DOI] [PubMed] [Google Scholar]
- 10. Roski F, Hammel J, Mei K, Baum T, Kirschke JS, Laugerette A, et al. Bone mineral density measurements derived from dual-layer spectral CT enable opportunistic screening for osteoporosis. Eur Radiol 2019; 29: 6355–63. doi: 10.1007/s00330-019-06263-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Kirkbride TE, Raja AY, Müller K, Bateman CJ, Becce F, Anderson NG. Discrimination between calcium hydroxyapatite and calcium oxalate using multienergy spectral photon-counting CT. AJR Am J Roentgenol 2017; 209: 1088–92. doi: 10.2214/AJR.17.18394 [DOI] [PubMed] [Google Scholar]
- 12. Li X, Li X, Li J, Jiao X, Jia X, Zhang X, et al. The accuracy of bone mineral density measurement using dual-energy spectral CT and quantitative CT: a comparative phantom study. Clin Radiol 2020; 75: 320. doi: 10.1016/j.crad.2019.11.008 [DOI] [PubMed] [Google Scholar]
- 13. Zhao L-Q, He W, Yan B, Wang H-Y, Wang J. The evaluation of haemodynamics in cirrhotic patients with spectral CT. Br J Radiol 2013; 86(1028. doi: 10.1259/bjr.20130228 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Dong Y, Zheng S, Machida H, Wang B, Liu A, Liu Y, et al. Differential diagnosis of osteoblastic metastases from bone islands in patients with lung cancer by single-source dual-energy CT: advantages of spectral CT imaging. Eur J Radiol 2015; 84: 901–7. doi: 10.1016/j.ejrad.2015.01.007 [DOI] [PubMed] [Google Scholar]
- 15. American College of Radiology . ACR practice guideline for the performance of quantitative computed tomography (QCT) bone densitometry. Resolution 33. 2008. Available from: http://www.acr.org/~/media/ACR/Documents/PGTS/guidelines/QCT.pdf
- 16. Hanley JA, McNeil BJ. A method of comparing the areas under receiver operating characteristic curves derived from the same cases. Radiology 1983; 148: 839–43. doi: 10.1148/radiology.148.3.6878708 [DOI] [PubMed] [Google Scholar]
- 17. Collier J. Bone disorders in chronic liver disease. Hepatology 2007; 46: 1271–78. doi: 10.1002/hep.21852 [DOI] [PubMed] [Google Scholar]
- 18. Ma Q, Hou X, Cheng X, You Y, Yang Z, Ma D, et al. Risk of vertebral fractures: evaluation on vertebral trabecular attenuation value and hydroxyapatite concentration in patients by chest spectral CT. Br J Radiol 2021; 94(1119. doi: 10.1259/bjr.20200234 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Baum T, Müller D, Dobritz M, Rummeny EJ, Link TM, Bauer JS. BMD measurements of the spine derived from sagittal reformations of contrast-enhanced MDCT without dedicated software. Eur J Radiol 2011; 80: e140–5. doi: 10.1016/j.ejrad.2010.08.034 [DOI] [PubMed] [Google Scholar]
- 20. Papadakis AE, Karantanas AH, Papadokostakis G, Petinellis E, Damilakis J. Can abdominal multi-detector CT diagnose spinal osteoporosis? Eur Radiol 2009; 19: 172–76. doi: 10.1007/s00330-008-1099-2 [DOI] [PubMed] [Google Scholar]
- 21. Griffith JF, Yeung DK, Antonio GE, et al. Vertebral marrow fat content and diffusion and perfusion indexes in women with varying bone density: MR evaluation. Radiology 2006; 241: 831–38. doi: 10.1148/radiol.2413051858 [DOI] [PubMed] [Google Scholar]
- 22. Ziemlewicz TJ, Maciejewski A, Binkley N, et al. Direct comparison of unenhanced and contrast-enhanced CT for opportunistic proximal femur bone mineral density measurement: implications for osteoporosis screening. AJR Am J Roentgenol 2016; 206: 694–98. doi: 10.2214/AJR.15.15128 [DOI] [PubMed] [Google Scholar]
- 23. Zopfs D, Lennartz S, Zaeske C, et al. Phantomless assessment of volumetric bone mineral density using virtual non-contrast images from spectral detector computed tomography. Br J Radiol 2020; 1: 1109. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Mei K, Schwaiger BJ, Kopp FK, Ehn S, et al. Bone mineral density measurements in vertebral specimens and phantoms using dual-layer spectral computed tomography. Sci Rep 2017; 7: 17519. [DOI] [PMC free article] [PubMed] [Google Scholar]