European Radiology Experimental | |
Comparison of volumetric and areal bone mineral density in CT and scout scans using spectral detector technology | |
Original Article | |
Jan Kirschke1  Marcus R. Makowski2  Lev Ushakov2  Daniela Pfeiffer2  Lorenz Birnbacher3  Johannes Hammel3  Franz Pfeiffer4  Graeme Campbell5  Philippe Coulon6  | |
[1] Department of Diagnostic and Interventional Neuroradiology, TUM School of Medicine, Klinikum Rechts Der Isar, Technical University of Munich, Munich, Germany;Department of Diagnostic and Interventional Radiology, TUM School of Medicine, Klinikum Rechts Der Isar, Technical University of Munich, Munich, Germany;Department of Diagnostic and Interventional Radiology, TUM School of Medicine, Klinikum Rechts Der Isar, Technical University of Munich, Munich, Germany;Chair of Biomedical Physics, Department of Physics, TUM School of Natural Sciences, Technical University of Munich, Garching, Germany;Department of Diagnostic and Interventional Radiology, TUM School of Medicine, Klinikum Rechts Der Isar, Technical University of Munich, Munich, Germany;Chair of Biomedical Physics, Department of Physics, TUM School of Natural Sciences, Technical University of Munich, Garching, Germany;Munich Institute of Biomedical Engineering, Technical University of Munich, Garching, Germany;TUM Institute for Advanced Study, Technical University of Munich, Garching, Germany;Philips GmbH Market DACH, Hamburg, Germany;Philips Healthcare, Suresnes, France; | |
关键词: Musculoskeletal diseases; Bone density; Tomography (x-ray computed); Osteoporosis; Absorptiometry (photon); | |
DOI : 10.1186/s41747-023-00356-7 | |
received in 2023-01-31, accepted in 2023-04-26, 发布年份 2023 | |
来源: Springer | |
【 摘 要 】
BackgroundTo determine whether denoised areal bone mineral density (BMD) measurements from scout scans in spectral detector computed tomography (CT) correlate with volumetric trabecular BMD for opportunistic osteoporosis screening.MethodsA 64-slice single-source dual-layer spectral CT scanner was used to acquire scout scan data of 228 lumbar vertebral bodies within 57 patients. Scout scans in anterior–posterior (AP) view were performed with a dose of < 0.06 mSv and spectrally decomposed into areal BMD (aBMD) values. A spectral dictionary denoising algorithm was applied to increase the signal-to-noise ratio (SNR). Volumetric trabecular bone mineral density (vBMD) was determined via material decomposition. A 3D convolutional network for image segmentation and labeling was applied for automated vBMD quantification. Projected maps were used to compare the classification accuracy of AP and lateral scout scans.ResultsThe denoising algorithm led to the minimization of anticorrelated noise in spectral maps and an SNR increase from 5.23 to 13.4 (p < 0.002). Correlation analysis between vBMD and measured AP aBMD, projected AP, and lateral aBMD showed a Pearson correlation coefficient of 0.68, 0.81, and 0.90, respectively. The sensitivity and specificity for the osteoporosis classification task were higher in lateral projection images than in AP crystallizing in an increased area under the curve value of 0.99 versus 0.90.ConclusionDenoised material-specific aBMD maps show a positive correlation to vBMD, enabling spectral scout scans as an opportunistic predictor for osteoporotic patients. This could be applied routinely as a screening tool in patients undergoing a CT examination.Relevance statementScout-based DEXA could be applied routinely as a screening tool in patients undergoing a CT examination.Key points• Spectral scout scans can be used as a dual-energy x-ray absorptiometry-like screening tool.• Spectral dictionary denoising on projection images increases the signal-to-noise ratio.• Positive correlation between volumetric and areal bone mineral density is observed.• Lateral projections increase osteoporosis classification accuracy compared to anterior-posterior projections.Graphical Abstract
【 授权许可】
CC BY
© The Author(s) 2023
【 预 览 】
Files | Size | Format | View |
---|---|---|---|
RO202309150136768ZK.pdf | 3352KB | download | |
13570_2023_282_Article_IEq17.gif | 1KB | Image | download |
Fig. 1 | 1656KB | Image | download |
13750_2023_309_Article_IEq64.gif | 1KB | Image | download |
Fig. 6 | 1327KB | Image | download |
Fig. 2 | 441KB | Image | download |
Fig. 3 | 353KB | Image | download |
MediaObjects/12974_2023_2872_MOESM3_ESM.docx | 3368KB | Other | download |
Fig. 6 | 857KB | Image | download |
MediaObjects/13750_2023_310_MOESM6_ESM.xlsx | 35KB | Other | download |
【 图 表 】
Fig. 6
Fig. 3
Fig. 2
Fig. 6
13750_2023_309_Article_IEq64.gif
Fig. 1
13570_2023_282_Article_IEq17.gif
【 参考文献 】
- [1]
- [2]
- [3]
- [4]
- [5]
- [6]
- [7]
- [8]
- [9]
- [10]
- [11]
- [12]
- [13]
- [14]
- [15]
- [16]
- [17]
- [18]
- [19]
- [20]
- [21]
- [22]
- [23]
- [24]
- [25]
- [26]
- [27]
- [28]
- [29]
- [30]
- [31]
- [32]
- [33]
- [34]
- [35]
- [36]
- [37]
- [38]
- [39]