| Background: Recently,the incidence of osteoporosis and osteopenia are increasing among Chinese.Osteoporosis is a metabolic bone disease that commonly seen in the elderly,cancer patients and postmenopausal women.It’s often neglected in daily life,which can easily lead to fractures and affect the quality of life.Quantitative computed tomography（QCT）is a data mining method based on the non-contrast CT scans,and it’s often used as a gold standard for the measurement of bone mineral density（BMD）.However,the specific platform and software of QCT are not yet generalized.Dual-energy CT（DECT）using the technology which can switch high and voltage instantaneous to provide more quantitative information,such as effective atomic number,calcium（water）density and energy spectrum curve,etc.In addition,the virtual non-contrast technology of DECT can removes the iodine in the contrast images to obtain virtual non-contrast scan（VNC）images that can replace the true non-contrast scan（TNC）images.VNC can also reduce the X-ray radiation dose while obtaining similar diagnostic performance,so it has a good application in the examinations of abdominal lesions.Objective:1.To investigate the feasibility of quantitative assessment of osteoporosis and osteopenia based on calcium-water separation technique of DECT with the BMD measured by QCT as the reference.2.Taking the BMD measured by QCT as the gold standard,to explore the feasibility of VNC of abdominal DECT to replace TNC,and the value of quantitative assessment of BMD and osteoporosis.1.160 volunteers who underwent abdominal DECT examinations were prospectively collected,and GE AW 4.6 workstation was used to generate the calcium-water map,and the calcium（water）density of L2 vertebra was measured.At the same time,the bone mineral density（BMD）of the L2 vertebra was measured using the QCT Pro workstation.The 160 patients were randomly divided into the test group and the validation group with 80 cases in each group.In the test group,the correlation between calcium（water）density and BMD of L2 vertebra was determined by Pearson correlation analysis,and Linear regression analysis was used to establish the regression model for calcium（water）density to predict BMD.The Bland-Altman method was used to analyze the consistency between the predicted BMD and the actual BMD in the validation group,and the discriminability and threshold of calcium（water）density in the diagnosis of osteoporosis and osteopenia were determined by receiver operating characteristic（ROC）curve.2.A retrospective analysis of 100 cancer patients who underwent abdominal non-contrast scan and three-phase contrast scan with gemstone spectral imaging（GSI）of DECT was performed.All patients’ age,height,weight and gender data were collected.The virtual non-contrast images were extracted from the GE AW 4.6workstation based on the venous contrast images,and the CT attenuation of L1 vertebra in VNC was measured.At the same time,the BMD of L1 vertebra were measured by QCT Pro software based on TNC images.The CT attenuation,standard deviation（SD）,signal-to-noise ratio（SNR）and contrast-to-noise ratio（CNR）of erector spinae were measured to compare the quality between VNC and TNC images.The dose-length product（DLP）and CT dose index（CTDI）of the four-phase scan were recorded.The 100 patients were randomly divided into training set and validation set with 50 patients in each set,and the correlation between the CT attenuation of VNC and BMD of 50 patients in the training set was determined by Pearson correlation analysis.Logistic multiple regression analysis was used to establish a regression model of patients age,height,weight,gender and the CT attenuation of VNC for predicting BMD.The accuracy of the model was verified by Bland-Altman consistency analysis in the validation set,and the Methods:performance of the CT attenuation of VNC in diagnosing osteoporosis and osteopenia was evaluated by ROC curve analysis.Results1.The calcium（water）density and BMD of L2 vertebra in the test group was62.99±17.26 mg/cm3 and 108.04±41.55 mg/cm3,respectively,which showed positive correlation（r=0.928,P<0.01）.The regression model of calcium（water）density and BMD was: BMD predict = 2.247×calcium（water）density-33.512（R2=0.928,P<0.05）.The calcium（water）density of L2 vertebra in the validation group was substituted into the model to obtain the predicted BMD.Bland-Altman analysis showed that there was no significant difference between the actual BMD and the predicted BMD of L2 vertebra in the validation group（P>0.05）.The ROC analysis showed that the threshold,AUC,sensitivity and specificity of calcium（water）density for the diagnosis of osteoporosis were 63.65 mg/cm3,0.908,98.43% and 75.41%,respectively;and 70.91 mg/cm3,0.825,84.44% and 77.14%for diagnosing osteoponia,respectively.2.The SNR and CNR of VNC images were slightly lower than those of TNC images（P<0.05）.There were no significant differences in the standard deviation（SD）of L1 vertebra,erector spinae and the CT attenuation of L1 vertebra between TNC and VNC images（all P>0.05）.The ED of TNC was（9.44±2.15）m Sv,and the DLP was（629.23±143.66）m Gy·cm;the effective dose（ED）of GSI mode venous contrast scan was（7.97±1.48）m Sv,and the dose length product（DLP）was（531.14±100.74）m Gy·cm,which compared with TNC decreased by 15.57%,and the total radiation dose of patients was reduced by 21.94% after using VNC technology.The CT attenuation of L1 vertebra of VNC was lower than that of TNC（CTTNC=133.57±48.42,CTVNC=71.24±26.38,P<0.05）,the CT attenuation of L1 vertebra in the VNC images of training set was significantly positively correlated with the BMD measured by QCT（r=0.974,P<0.05）.The multiple linear regression analysis showed that the BMD was correlated with the patients age,height and the CT attenuation of VNC.The regression model was established as follows:BMD=102.375+1.349 × CT attenuation-0.321 × age-0.433 × height（R2=0.979,P<0.05）.The patients’ age,height and the CT attenuation of VNC in the validation set were brought into the model to obtain the predicted BMD.There was no statistical difference between the predicted BMD and the actual BMD measured by QCT（P>0.05）.The ROC analysis showed that the cut-off value of the CT attenuation of VNC for the diagnosis of osteopenia was 64.51 HU（AUC: 0.964,95%CI: 0.922-0.997）,and the cut-off value for the diagnosis of osteoporosis was54.35 HU（AUC: 0.976,95%CI: 0.913-0.991）.Conclusion1.The calcium（water）separation technology of DECT can quantitatively diagnose osteoporosis and osteopenia,which providing a new ideal for opportunistic screening of osteoporosis without phantom.2.The VNC images of DECT is expected to replace conventional non-contrast scan which also can to reduce patien radiation dose,and it’s significant for assessing bone mineral density changes and osteoporosis during the follow-up of cancer patients. |
PDF Full Text Request