| Objective:To explore the feasibility of the material decomposition and quantification technique of gemstone spectral CT for measuring bone mineral density(BMD).Methods:A total of 210 patients who admitted to the hospital and underwent abdominal examination with gemstone spectral CT from July 2017 to September 2018 were selected as the research group.Among them,104 cases were males,aged from 17 to 82 years old,with an average age of 47.50±16.07 years old;106 cases were females,aged from 16 to 87 years old,with an average age of 48.88±16.59 years old.All the male and female patients were divided into 6 groups according to their ages,respectively<30,30～39,40～49,50～59,60～69 and≥70 years old groups.At the same time,301 patients who underwent dual-energy X-ray absorptiometry(DXA)examination with the same background and period were selected as the control group with the same group division,of which 149 cases were males,aged from 24 to 86 years old,with an average age of 50.80±16.34 years old;152 cases were females,aged from 20 to 86 years old,with an average age of 51.22±16.10 years old.All cases were analyzed by the software of gemstone spectral imaging in the resesrch group.Three regions of interest(ROI)were placed at the central slice and the upper and lower slices on the axial view of L1,L2,L3 and L4 vertebral bodies of the patient’s images respectively,the average values of 12 measurement results were obtained,then the mean densities of calcium(water),hydroxyapatite(water)of L1-L4 vertebral bodies were obtained.Mean BMD value of L1-L4 vertebral bodies was obtained in the control group.Correlation coefficients were used to analyze the correlation between mean densities of calcium(water),hydroxyapatite(water),mean BMD of L1-L4 vertebral bodies and age,and the correlation between the mean calcium(water),hydroxyapatite(water)densities and mean BMD(Pearson correlation analysis was performed when the data were subject to normal distribution,and Spearman correlation analysis was performed when the data were not subject to normal distribution).Results:The peak values of mean densities of calcium(water)and hydroxyapatite(water)of L1-L4 vertebral bodies and mean BMD of L1-L4 vertebral bodies obtained by DXA were in the group of<30 years old in males,and in the group of 30～39 years old in females.The mean densities of calcium(water)and hydroxyapatite(water)in males were negatively correlated with age(r=-0.619 and-0.619,respectively,both P<0.001).The mean BMD of L1-L4 vertebral bodies obtained by DXA also decreased gradually with age,but increased when patients were>70 years old.Correlation analysis showed that BMD was negatively correlated with age(r=-0.351,P<0.001).The mean densities of calcium(water),hydroxyapatite(water)of L1-L4 vertebral bodies and mean BMD of L1-L4 vertebral bodies obtained by DXA in females were negatively correlated with age(r=-0.689,-0.690 and0.606,respectively,all P<0.001).Among them,calcium(water)density,hydroxyapatite(water)density and BMD showed no correlation with age when patients were<40 years old(all P>0.05),while they were negatively correlated with age when patients were>40 years old(r=-0.671,-0.674 and-0.561,respectively,all P<0.001).The mean densities of calcium(water)and hydroxyapatite(water)of L1-L4 vertebral bodies in males and females were positively correlated with the mean BMD of L1-L4 vertebral bodies obtained by DXA(r=0.714,0.714 and 0.720,0.720,respectively,all P<0.001).Conclusion:The densities of calcium(water)and hydroxyapatite(water)of lumbar vertebra were positively correlated with the BMD measured by DXA,and showed the same change trend with age,which indicated that the densities of calcium(water)and hydroxyapatite(water)could reflect the changes of bone content,and the material decomposition and quantification technique of gemstone spectral CT can be used to measure the BMD of lumbar vertebrae.Objective:To explore the value of the material decomposition and quantification technique of gemstone spectral CT in the diagnosis of osteoporosis(OP).Methods:The imaging data of 223 patients who underwent abdominal examination with gemstone spectral CT and dual-energy X-ray absorptiometry(DXA)in our hospital from September 2018 to August 2020 were analyzed retrospectively.Among them,92 cases were males,aged from 53 to 94 years old,with an average age of 72.64±9.72 years old;101 cases were females,aged from 16 to 87 years old,with an average age of 72.27±9.70 years old.Three regions of interest(ROI)were placed at the central slice and the upper and lower slices on the axial view of L1,L2,L3 and L4 vertebral bodies of the patient’s CT images with the software of GSI,the average values of 12 measurement results were obtained,then the mean densities of bone(fat),bone(water),calcium(fat),calcium(water),hydroxyapatite(fat),hydroxyapatite(water),iron(fat),iron(water),fat(water),and water(fat)of L1-L4 vertebral bodies were obtained.DXA performed as gold standard(The T score calculated from the average bone mineral density of L1-L4 vertebral bodies was taken,and the T score≤-2.5SD was considered as osteoporosis),the male and female patients were divided into non-osteoporosis group and osteoporosis group.Independent sample t-test was used to compare the difference of values of spectral CT parameters between two groups.Correlation coefficients were used to analyze the correlation between the values of CT parameters and BMD measured by DXA(Pearson correlation analysis was performed when the data were subject to normal distribution,and Spearman correlation analysis was performed when the data were not subject to normal distribution).Unary linear regression analysis was used to compare the correlation between the values of spectral CT parameters and BMD measured by DXA.The receiver Operating characteristic(ROC)curve was used to evaluate the diagnostic efficiency of all parameters of spectral CT in osteoporosis,and the Z test was used to compare the areas under the curve(AUC)of different parameters.Then the parameter with the optimal diagnostic efficiency of osteoporosis,and the optimal critical value of this parameter in the diagnosis of osteoporosis were obtained.Results:All values of spectral CT parameters in the non-osteoporosis group and osteoporosis group were statistically different in males and females(P<0.05).The mean density values of fat(water)of L1-L4 vertebral bodies in males and females were negatively correlated with mean BMD of L1-L4 vertebral bodies measured by DXA,while the values of other parameters were positively correlated with BMD(P<0.001).Among them,the correlation between hydroxyapatite(water)density and BMD was the highest(r=0.827 and 0.831,respectively,both P<0.001).Unary linear regression analysis showed that the values of spectral CT parameters in males and females were linearly related to the BMD measured by DXA,among which the regression model of hydroxyapatite(water)density and BMD had the best fitting degree(R2=0.680 and 0.689,respectively,both P<0.001).ROC curves showed that all parameters of spectral CT in males and females had diagnostic efficacy for osteoporosis.The comparison of area under the curve(AUC)showed that there was no statistical difference of all spectral CT parameters(all P>0.05).But among them,the AUC value of hydroxyapatite(water)for the diagnosis of osteoporosis was the largest(AUC=0.852 and 0.890,respectively,both P<0.001).When the mean hydroxyapatite(water)density of L1-L4 vertebral bodies in males was 45.470mg/cm3 for the diagnosis of osteoporosis,the diagnostic sensitivity,specificity,positive predictive value and negative predictive value were 82.5%,80.8%,0.767 and 0.857,respectively.When the mean hydroxyapatite(water)density of L1-L4 vertebral bodies in females was 53.650mg/cm3 for the diagnosis of osteoporosis,the diagnostic sensitivity,specificity,positive predictive value and negative predictive value were 81.4%,83.6%,0.851 and 0.797,respectively.Conclusion:The material decomposition and quantification technique of gemstone spectral CT can be used in the diagnosis of osteoporosis.Hydroxyapatite(water)is the best base material pair for diagnosis of osteoporosis. |
PDF Full Text Request