KR101750108B1

KR101750108B1 - Estimation of Bone Mineral Density Using Medical Image Signal Without an External Phantom

Info

Publication number: KR101750108B1
Application number: KR1020150107946A
Authority: KR
Inventors: 이영한
Original assignee: 연세대학교 산학협력단
Priority date: 2015-07-30
Filing date: 2015-07-30
Publication date: 2017-06-26
Also published as: KR20170015674A

Abstract

The present invention includes the step of obtaining a bone mineral density (BMD) using an HU (Hounsfield unit) value or (b) HU (Hounsfield unit) value and a patient specific factor without an external phantom And a method for measuring bone density. Using the regression equation for HU to BMD conversion of the present invention, it is possible to reliably calculate BMD values from CT images without external phantoms, thereby avoiding additional radiation dose due to QCT testing for BMD measurements. In addition, the model of the present invention can be used as a feasible tool for FEA-based osteoporosis research using generic CT images for large populations. In addition, the present invention can calculate the bone mineral density using a general medical image such as MRI or CT without an external phantom through calculation formula in each patient based on the above concept. In addition, the calculated bone density can be used to predict bone strength based on bone density, and can be used for patient-specific diagnosis or treatment.

Description

Estimation of Bone Mineral Density Using Medical Image Signal Without an External Phantom

The present invention includes the step of obtaining a bone mineral density (BMD) using an HU (Hounsfield unit) value or (b) HU (Hounsfield unit) value and a patient specific factor without an external phantom The present invention relates to a method for measuring bone density.

Osteoporosis is a common metabolic bone disease that increases the risk of fracture. As the elderly population increases, the prevalence of osteoporosis is also expected to increase (1, 2); Therefore, calculation of bone strength is becoming more important as a diagnostic tool for osteoporosis. Essentially, bone strength depends on two parameters: bone quality (eg, bone structure) and amount of bone (eg, bone density). In clinical trials, the calculation of bone strength is based on a representative areal bone mineral density (aBMD) for the desired anatomic location (eg, femur, cuff, and vertebra) obtained using dual energy X-ray absorptiometry (3-5). However, due to the unique 2D projection that ignores out-of-plane changes in the bone structure, aBMD alone could not measure the bone strength completely. It has also been reported that two individuals with the same aBMD may have different bone structures and consequently have different fracture risk (6).

Volumetric BMDs (vBMDs), as contrasted with aBMD, are measured using quantitative computed tomography (CT) (7, 8), which can provide a BMD distribution in 3D space, The sources of errors can be eliminated. For the lumbar spine and proximal femur, where trabecular bone is widely distributed, QCT can provide greater diagnostic sensitivity for bone strength calculations compared to DXA (9). QCT also excludes osteophytes 10 and aortic / vascular calcifications 9, which can affect BMD. To accurately calculate the BMD values, QCT requires a reference phantom constructed from K ₂ HPO ₄ with a known density value (11, 12). The reference phantom removes potential confounding factors from scattering and beam hardening, which depend on individual patient factors such as waist circumference and body weight.

In clinical trials routine CT (computed tomography) scans (ie, phantomless CT scans) are obtained for a variety of purposes. It would be useful to calculate BMD values using these general CT scans. Up to now, several HU-to-BMD transformations have been introduced in BMDs of hounsfield units (13-16). The various effects of the scanning protocol on HU values, including contrast mediums 17,18 and kVp 19, have been further investigated since BMD values are significantly related to HU values. It should be noted that patient parameters such as waist circumference and bone area should be considered for accurate conversion of HU to BMD because these factors can affect radiation attenuation and affect HU values. However, until now there has been no literature dealing with the results of HU-to-BMD conversion of patient parameters.

Numerous papers and patent documents are referenced and cited throughout this specification. The disclosures of the cited papers and patent documents are incorporated herein by reference in their entirety to better understand the state of the art to which the present invention pertains and the content of the present invention.

The present inventors have sought to develop a method for predicting bone density through general CT rather than quantitative computed tomography (QCT) for measuring bone mineral density (BMD). As a result, we propose a regression model for the conversion of phantom-free HUs containing patient-related factors into BMD (HU-to-BMD), compare the BMD values derived with the reference phantom to the predicted BMD values, The BMD value, which is the most important marker of the bone strength, can be easily calculated by a simple equation using only the HU value obtained by the general CT, thereby completing the present invention.

Accordingly, an object of the present invention is to provide a method of measuring bone mineral density (BMD) using a Houns field unit (HU) value and / or a patient specific factor without an external phantom .

Other objects and advantages of the present invention will become more apparent from the following detailed description of the invention, claims and drawings.

According to one aspect of the present invention, the present invention relates to a method and apparatus for evaluating a human body, comprising the steps of: (a) determining a Hounsfield unit (HU) value or (b) Hounsfield unit value and a body weight, And determining a bone mineral density (BMD) using a patient specific factor selected from the group consisting of: < RTI ID = 0.0 >

The present inventors have sought to develop a method for predicting bone density through general CT rather than quantitative computed tomography (QCT) for measuring bone mineral density (BMD). As a result, we propose a regression model for the conversion of phantom-free HUs containing patient-related factors into BMD (HU-to-BMD), compare the BMD values derived with the reference phantom to the predicted BMD values, It was confirmed that the BMD value, which is the most important marker of the bone strength, can be easily calculated by a simple equation using the HU value obtained from the general CT.

According to one embodiment of the present invention, the bone mineral density measured by the method of the present invention is the bone density of bone included in CT images such as lumbar spine, femur, humerus, radius, knee or ankle, and most preferably, to be.

According to another embodiment of the present invention, the method includes obtaining a lumbar spine's bone density using an HU value and a waist circumference without an external phantom, more preferably, the method comprises measuring the lumbar spine (Hounsfield unit) value and circumference of the body of a lumbar vertebra (BM) according to Equation (1) to obtain a bone mineral density (BMD) of the lumbar vertebra:

Equation 1

( BMD ) _{lumbar spine} = 0.89 占 ( HU ) _{lumbar spine} + 0.15 占 (waist circumference) -7.44

In Equation (1), the unit of ( BMD ) _{lumbar spine} is mg / cc, and the unit of the waist circumference is cm.

According to another embodiment of the present invention, the method comprises determining the bone density of the hip using only the HU value without an external phantom, more preferably, the method comprises determining the HU Hounsfield unit) value to the following equation (2) to obtain the bone mineral density (BMD) of the femur:

Equation 2

( BMD ) _hip = 0.78 x ( HU ) _hip + 16.62

In Equation (2), the unit of ( BMD ) _hip is mg / cc.

According to the present invention, QCT images taken using a reference phantom are retrospectively analyzed for L2 vertebrae and proximal femurs, and HU values recorded including the HU values of reference phantoms are recorded Then, univariate analysis of patient-related data, such as waist circumference (cm) and bone cross-sectional area (cm ² ), HU values and BMD values were performed and only statistically significant factors were included in the multivariate analysis, A multiple linear regression model was used to convert the phantom-free HU to BMD.

And, for statistical analysis, the correlation between predicted BMD values (i.e., phantom-free data) and reference BMD values (i.e., phantom-based data) was evaluated using Pearson correlation tests, For further application, voxelwise comparison was performed using root mean square error (RMSE).

As a result, in the univariate analysis, the HU values and circumference were statistically significant ( p <0.05) for the lumbar spine and only the HU values for the proximal femur were statistically significant ( p <0.05) From the regression model, the conversion equation of the phantom-free HU to the BMD for the lumbar and proximal femur, that is, the above equations 1 and 2, was established.

The predicted BMD values correlated significantly with the BMD values measured using the reference phantom. In the voxelwise comparison, the RSME values of the lumbar spine and femur were confirmed.

Thus, the present invention derives a transform equation that includes the perimeter and bony area, examines the correlation between the predicted BMD values and the BMD values obtained using the reference phantom, and provides the proposed model for finite element analysis And the voxelwise accuracy of the test was analyzed.

According to a more specific embodiment of the present invention, the HU value of the lumbar spine or the HU value of the femur is obtained from CT (computed tomography) or MRI (magnetic resonance imaging) images without external phantom, It is obtained from a CT (computed tomography) image without an external phantom.

According to another more specific embodiment of the present invention, the reference bone density for comparison with the bone density measured from Equation (1) or (2) is defined by Equation (3)

Equation 3

( BMD ) _phantom is a value measured using quantitative computed tomography ( CTT ) with an external phantom, HU is a value corresponding to an external phantom, BMD of the lumbar vertebrae measured from the above and the Pearson correlation coefficient (Pearson's correlation coefficient) in relation to (BMD) of the equation (3) _phantom 0.986, bone density of the femur determined from the equation (2) is (BMD of the equation (3) ) The Pearson correlation coefficient is 0.948 in relation to _phantom .

The α and β in equation 3 is the patient to be determined using the HU value corresponding to the bone density and outside phantom measured by QCT taking-a specific value.

According to another embodiment of the present invention, the bone density measured from Equation (1) or Equation (2) and the reference bone density of Equation (3) may be expressed as a root mean square error mean square error (RMSE):

Equation 4

( BMD _phantom ) _i is the BMD value of the i- th voxel calculated using Equation (3), and BMD _phantomless _i is calculated using Equation (1) or Equation (2) a i is the BMD value of the first voxel (voxel), wherein n is the total number of voxels (voxel), RMSE for the lumbar spine is 4.26 ± 0.60 [mg / cc] and the RMSE for the femur is 8.35 ± 0.57 [mg / cc].

According to a more specific embodiment of the present invention, the quantitative CT is performed under conditions of 120 kVp, effective mAs 150 mAs, slice thickness of 3 mm and B40s (medium) kernel And most preferably 120 kVp, effective mAs 150 mAs, beam collimation = 20 mm, rotation speed = 0.6 sec, pitch = 1.0: 1, 512 x 512 matrix, field of view (FOV) = 360 mm, slice thickness = 3 mm, and reconstruction with B40s (medium).

The features and advantages of the present invention are summarized as follows:

(I) The present invention relates to a method for obtaining a bone mineral density (BMD) using an HU (Hounsfield unit) value or (b) HU (Hounsfield unit) value and a patient specific factor without an external phantom A method for measuring bone mineral density comprising the steps of:

(Ii) Using the regression equation for converting the HU of the present invention to BMD, it is possible to reliably calculate BMD values from CT images without external phantoms, thereby avoiding additional radiation dose due to QCT test for BMD measurement .

(Iii) In addition, the model of the present invention can be used as a feasible tool for FEA-based osteoporosis research using generic CT images for large populations.

(Iv) Based on the above concept, the present invention can calculate the bone density using a general medical image such as MRI or CT without an external phantom through a calculation formula in each patient.

(V) And, the computed bone density images can be used to diagnose or treat the bone strength of bone-density-based bone specimens.

Figures 1a and 1b show a screen shot of semi-automatic calculation software (FatScan, N2 systems). FIG. 1A shows that the circumference of the trunk and the cross-sectional area of the bone are divided at the L2 level, and FIG. 1B shows that the cross-sectional area of the circumference and bone of the trunk is divided at the level of the femur.
2a and 2b show Pearson correlation tests of BMD values derived using predicted BMD values and a reference phantom. FIG. 2A shows the Pearson correlation coefficient of 0.986 for the L2 level, and FIG. 2B shows the Pearson correlation coefficient of 0.948 for the femoral level ( p < 0.05).
Figure 3 shows a BMD contour plot at the axial level of the L2 vertebra. Panel (a) shows the BMD deviation, panel (b) shows the predicted BMD, panel (c) shows the BMD deviation, and panel (d) shows the BMD deviation with another legend.
Figure 4 shows a BMD contour plot in the axial direction of the hip joint. Panel (a) shows the BMD deviation, panel (b) shows the predicted BMD, panel (c) shows the BMD deviation, and panel (d) shows the BMD deviation with another legend.

Hereinafter, the present invention will be described in more detail with reference to Examples. It is to be understood by those skilled in the art that these embodiments are only for describing the present invention in more detail and that the scope of the present invention is not limited by these embodiments in accordance with the gist of the present invention .

Example

Materials and Methods

Study population

The research population was retrospectively verified using the hospital information system. The inclusion criteria were based on (1) the examination conducted in April 2014, and (2) no bony abnormalities on the radiology report. For 39 confirmed cases, the purpose of the QCT test was physical examination (n = 36), breast cancer follow-up (n = 1) and thyroid cancer follow-up (n = 2). Sex distribution was 14 males and 25 females. The mean age was 49.1 years (range: 30-73 years). This retrospective study was approved by the Institutional Review Board (IRB) of the hospital.

Image protocol

QCT scans were performed on 64-channel CT (Somatom Definition AS +, Siemens, Erlangen, Germany). CT scan parameters were optimized for the QCT test as follows: 120 kVp, effective mAs 150 mAs, beam collimation = 20 mm, rotation speed = 0.6 s, pitch, = 1.0: 1, 512 x 512 matrix, field of view (FOV) = 360 mm, slice thickness = 3 mm, and reconstruction with B40s (medium). CARE Dose 4D and CARE kV did not work. The CT dose index (CTDI 32 cm) was 10.10 mGy and the DLP (dose length product) was 218.7 mGy · cm. A deep-inspiration breath-hold (DIBH) method was used for QCT scans.

Calculation of reference BMD using external phantom

Five regions of interest (ROIs) for five different mineral contents of an external phantom (Mindways Inc., Austin, Tex., USA) were recorded on the same CT image. For each CT image of the patients, the phantom-based calibration algorithm 20, 21 determined a linear correlation between the known bone density, defined by the following equation (1), and the corresponding HU value:

(One)

Here ,? And ? Are patient-specific values determined. Then, using the formula (1) having an α and β is determined, the value of the lumbar BMD (lumbar spines) and femur (hips) were calculated as the standard (reference) for comparison.

Image analysis

All images were evaluated by a radiologist who had 10 years of experience in musculoskeletal radiology and completed a course in musculoskeletal medicine. 39 QCT images of other patients were analyzed retrospectively in L2 vertebra and axial levels of the femur. Quantitative evaluation of the ROI was performed on 80-100 mm ² drawings of the trabecular compartment of the L2 body and the entire femur in which HU values were recorded.

Calculation of the circumference of the trunk and the bone cross-sectional area

As shown in Fig. 1, the circumference of the trunk and the bone cross-sectional area were measured in the same axial direction of the target site using semiautomatic calculation software (FatScan, N2 systems, Osaka, Japan). Because the software can calculate subcutaneous and visceral fat areas and their proportions, they were used to calculate the circumference and bone cross-sectional area of the trunk.

Regression models for conversion of HU-to-BMD to phantom-free HU BMD and their correlation test

Circumference, bone area, HU values and BMD values were considered for multivariate analysis. Using this data, a multivariate regression model was used as a step - by - step regression method to establish a phantom - free HU to BMD conversion equation. The independent variable was the BMD value, and the dependent variable was the HU value, circumference, and bone area.

Next, a Pearson correlation test was performed to investigate the correlation between the predicted BMD values and the predicted values in Equation (1). All statistical analyzes were performed using statistical software (R package 2.15.1; http://cran.r-project.org). P -values less than 0.05 were considered statistically significant.

Comparison of voxelwise BMD

For voxelwise comparison, the HU values of lumbar and femur in the same axial direction were reconstructed into 512x512 array data. Then, the predicted BMD and the reference BMD were voxelly calculated using the proposed model (Equation 3) and Equation 1, respectively. They were then compared using the root mean square error (RMSE) of Equation 2:

(2)

Here, BMD _phantom _i and BMD _phantomless _i are the BMD values of the i- th voxel calculated using Equation 1 and BMD values of the i- th voxel estimated using the proposed model, Value; n Represents the total number of voxels. Conversion and statistical calculations were performed using commercially purchased software (Interactive Data Language (IDL), Exelis Vis Inc., Boulder, CO, USA).

Experiment result

For each patient's CT image, α and β in Equation 1 were determined using known bone density and corresponding HU values of the external phantom. BMD values of the lumbar and femur were then calculated as reference data.

From univariate analysis, it was determined that the bone area was not significant for both lumbar spine and femur ( p > 0.05). On the other hand, HU values and trunk circumference were statistically significant ( p <0.05) for the lumbar spine, and only HU values were statistically significant for the femur ( p <0.05). The HU-to-BMD conversion equation was then established from the multiple linear regression model, taking into account the patient factors in the lumbar spine and femur as follows:

(3)

Here, the units of BMD and perimeter are mg / cc and cm, respectively.

From the Pearson correlation test (FIG. 2), it was found that the BMD values predicted from Equation 3 were significantly correlated with the reference BMD values from Equation 1 (Pearson correlation coefficients for lumbar and femur were 0.986 and 0.948; p < 0.05). In voxelwise comparisons, RSME values of lumbar spine and femur were 4.26 ± 0.60 [mg / cc] and 8.35 ± 0.57 [mg / cc], respectively.

Review

Because bone mineral density (BMD) is an important marker of bone strength (22), it plays an important role in the diagnostic criteria and treatment response to osteoporosis. However, routine CT scans (ie, external phantomless CT scans) do not provide BMD values directly, but rather provide HU values. For accurate BMD measurements, external solid phantoms were placed in a CT scanner to compensate for the effects of beam hardening and radiation scattering (11, 12).

In the prior art, HU and BMD values have been reported to have a linear correlation (23), and some regression models include contrast-enhanced CT (13,14), CT colonography (17,24) An abdominal multi-detector CT 15, and a spine CT 16, 25 have been proposed. These studies have shown that the correlation between HU and BMD values is dependent on the contrast medium, kVp, and CT scanning regions. Although the HU values are affected by patient factors such as waist circumference and cross-sectional area of the bone (11, 26), there is no literature dealing with patient factors and BMD calculations.

The present invention has been developed by the need to reliably calculate the BMD without external phantom taking into account the effects on the HU values of the patient parameters. With a total of 39 QCT images, the relationship between HU values and patient parameters was investigated through multivariate analyzes, resulting in a multiple linear regression model for the phantom-free HU to BMD conversion equation. The high positive correlation between the predicted BMD values and the reference BMD values is shown through Pearson correlation tests. From the results of the present invention, the mean BMD values of the lumbar spine using the HU values and the perimeter equations could be calculated, but the mean BMD values of the femur could be calculated using only the HU values. The cross-sectional area of the bone was determined not to be significant in both lumbar spine and femur. Thus, the proposed transformations can statistically compensate for reduced HU values due to patient factors.

Figures 3 and 4 clearly show the similarity of the predicted BMD and reference BMD of each voxel. These deviations can be regarded as negligible (i.e., 1000 mg / cc in FIG. 3 (c) and FIG. 4 (c)) compared with the maximum BMD value.

For thorough verification, the BMD values predicted by the RMSE equation, which can provide a representative value for the BMD deviation of each voxel, were compared with the reference BMD values. Considering that the RMSE for lumbar spine and femur is 4.26 ± 0.60 [mg / cc] and 8.35 ± 0.57 [mg / cc] respectively, the proposed models can provide reliable BMD values for each voxel, BMD distribution. It should be noted that the BMD distribution data is essential for constructing FE (finite element) models. According to Crawford et al . (27), FE models can be a more reliable tool for fracture risk assessments. Also, as can be seen in Figures 3 (d) and 4 (d), the legend of maximum BMD for more clear visualization is different from Figures 3 (c) and 4 It is also interesting to note that BMD deviations are proportional to their BMD values. These errors are proportional to occur from a predetermined slope and y- intercept statistically in formula 3, which attenuation (attenuation) to the patient - to compensate for patient-specific differences - the specific values (that is, the formula 1 α and β )to be.

Looking at the constraints of the present invention, the proposed HU to BMD conversion equation is based on the QCT protocol: 120 kVp, effective mAs 150 mAs, slice thickness of 3 mm, and B40s ( medium) The kernel. For better accessibility, the proposed transform equation is extended to include a generic CT protocol. However, considering the radiation dose modulation techniques, simple expressions of kVp or mAs may be inadequate. In addition, the effect of increased density of the contrast agent was not evaluated. Since CT density varies with phase after contrast injection, phantom-free BMD calculations should be performed with caution when using contrast-enhanced CT images.

As a follow-up study, the proposed model for the conversion of phantom-free HU to BMD (HU-to-BMD) can be extended to using generic CT images. Since a general CT scan is performed on a daily practice basis, the proposed model will allow patients to avoid additional radiation exposure for BMD measurements. In addition, with the help of commercially available Picture Archiving and Communication System (PACS), FEA-based fracture risk assessment can be used to study osteoporosis with large populations, which is more accurate for translational medicine Meaningful diagnostic data can be provided.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the same is by way of illustration and example only and is not to be construed as limiting the scope of the present invention. Accordingly, the actual scope of the present invention will be defined by the appended claims and their equivalents.

references

1. Iki M. Epidemiology of bone and joint disease - the present and future. Epidemiology of osteoporosis and osteoporotic fracture in Japan. Clinical calcium. 2014; 24: 657-664.

2. Leslie WD, Morin SN. Osteoporosis epidemiology 2013: implications for diagnosis, risk assessment, and treatment. Current opinion in rheumatology. 2014; 26: 440-446.

3. Blake GM, Fogelman I. The role of DXA bone density scans in the diagnosis and treatment of osteoporosis. Postgraduate medical journal. 2007; 83: 509-517.

4. Finkelstein JS, Klibanskie, Neer RM. Evaluation of lumbar spine bone mineral density (BMD) using dual energy x-ray absorptiometry (DXA) in 21 young men with histories of constitutionally-delayed puberty. The Journal of Clinical Endocrinology and Metabolism. 1999; 84: 3400-3401; author reply 3403-3404.

5. Pereira RM, Corrente JE, Chahade WH, Yoshinari NH. Evaluation of dual X-ray absorptiometry (DXA) of bone mineral density in children with juvenile chronic arthritis. Clinical and experimental rheumatology. 1998; 16: 495-501.

6. Marshall D, Johnell O, and Wedel H. Meta-analysis of how well measures of bone mineral density predicted osteoporotic fractures. Bmj. 1996; 312: 1254-1259.

7. ACRSPRSSR practice guideline for quantitative computed tomography (QCT) bone densitometry available at http://www.acr.org/~/media/DE78D218C7A64526A821A9E8645AB46D.pdf. Accessed April 25, 2014. Accessed.

8. Lewiecki EM. Imaging technologies for assessment of skeletal health in men. Current osteoporosis reports. 2013; 11: 1-10.

9. Li N, Li XM, Xu L, Sun WJ, Cheng XG, Tian W. Comparison of QCT and DXA: Osteoporosis Detection Rates in Postmenopausal Women. International journal of endocrinology. 2013; 2013: 895474.

10. Ito M, Hayashi K, Yamada M, Uetani M, Nakamura T. Relationship of osteophytes to bone mineral density and spinal fracture in men. Radiology. 1993; 189: 497-502.

11. Adams JE. Quantitative computed tomography. European journal of radiology. 2009; 71: 415-424.

12. Habashy AH, Yan X, Brown JK, Xiong X, Kaste SC. Estimation of bone mineral density in children from diagnostic CT images: a comparison of methods with and without an internal calibration standard. Bone. 2011; 48: 1087-1094.

13. Bauer JS, Henning TD, Mueller D, Lu Y, Majumdar S, Link TM. Volumetric quantitative CT of the spine and hip derived from contrast-enhanced MDCT: conversion factors. AJR American journal of roentgenology. 2007; 188: 1294-1301.

14. Baum T, Muller D, Dobritz M, et al. Converted lumbar BMD values derived from sagittal reformations of contrast-enhanced MDCT predict incidental osteoporotic vertebral fractures. Calcified tissue international. 2012; 90: 481-487.

15. Papadakis AE, Karantanas AH, Papadocostakis G, Petinellis E, Damilakis J. Can abdominal multi-detector CT diagnose spinal osteoporosis? European radiology. 2009; 19: 172-176.

16. Schreiber JJ, Anderson PA, Rosas HG, Buchholz AL, Au AG. Hounsfield units for assessing bone mineral density and strength: a tool for osteoporosis management. The Journal of the American Medical Association. 2011; 93: 1057-1063.

17. Fletcher JG, Johnson CD, Krueger WR, et al. Contrast-enhanced CT colonography in recurrent colorectal carcinoma: feasibility of simultaneous evaluation for metastatic disease, local recurrence, and metachronous neoplasia in colorectal carcinoma. AJR American journal of roentgenology. 2002; 178: 283-290.

18. Baum T, Muller 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. European journal of radiology. 2011; 80: e140-145.

19. Schwaiger BJ, Gersing AS, Baum T, Noel PB, Zimmer C, Bauer JS. Bone mineral density values derived from routine lumbar spine multidetector row CT predicted osteoporotic vertebral fractures and screw loosening. AJNR American journal of neuroradiology. 2014; 35: 1628-1633.

20. Emamia, Ghadiri H, Ay MR, et al. A new phantom for performance evaluation of bone mineral densitometry using DEXA and QCT. Book A new phantom for performance evaluation of bone mineral densitometry using DEXA and QCT. City2011; 3441-3445.

21. Cetin Celenk and Peruze Celenk (2012). Bone Density Measurement Using Computed Tomography, pp. 131, Computed Tomography - Clinical Applications, Luca Saba (Ed.), ISBN: 978-953-307-378-1, InTech, DOI: 10.5772 / 22884. Available from: http://www.intechopen.com/books/computed-tomography-clinical-applications/bone-density-measurement-using-computed-tomography.

22. Lenchik L, Shi R, Register TC, Beck SR, Langefeld CD, Carr JJ. Measurement of trabecular bone mineral density in the thoracic spine using cardiac gated quantitative computed tomography. Journal of computer assisted tomography. 2004; 28: 134-139.

23. 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: 805-811.

24. Summers RM, Baecher N, Yao J, et al. Feasibility of simultaneous computed tomographic colonography and fully automated bone mineral densitometry in a single examination. Journal of computer assisted tomography. 2011; 35: 212-216.

25. Schwaiger BJ, Gersing AS, Baum T, Noel PB, Zimmer C, Bauer JS. Bone Mineral Density Values Derived from Routine Lumbar Spine Multidetector Row CT Predict Osteoporotic Vertebral Fractures and Screw Loosening. AJNR American journal of neuroradiology. 2014.

26. Celenk C, Celenk P. Bone density measurement using computed tomography. Book Bone density measurement using computed tomography. City2012; pp. 131.

27. Crawford RP, Cann CE, Keaveny TM. Finite element models predict in vitro vertebral body compressive strength better than quantitative computed tomography. Bone. 2003; 33: 744-750.

Claims

Using patient specific factors selected from the group consisting of (a) Hounsfield unit (HU) value or (b) Hounsfield unit value and body weight, height and circumference of body without external phantom To obtain a bone mineral density (BMD).

The method according to claim 1, wherein the bone density is a bone density of the lumbar region, the femur, the humerus, the radius, the knee, or the ankle joint.

2. The method of claim 1, wherein the method comprises determining a bone density of a lumbar spine using an HU value and a waist circumference without an external phantom.

2. The method of claim 1, wherein the method comprises determining a bone density of the hip using only the HU value without an external phantom.

4. The method according to claim 3, wherein the method comprises: obtaining a bone mineral density (BMD) of a lumbar spine by substituting a Hounsfield unit (HU) value of a lumbar spine and a circumference of a body into Equation Comprising the steps < RTI ID = 0.0 > of:
Equation 1
( BMD ) _{lumbar spine} = 0.89 占 ( HU ) _{lumbar spine} + 0.15 占 (waist circumference) -7.44
In Equation (1), the unit of ( BMD ) _{lumbar spine} is mg / cc, and the unit of the waist circumference is cm.

5. The method according to claim 4, wherein the method comprises: obtaining a bone mineral density (BMD) of a femur by substituting a Hounsfield unit (HU) value of the hip into the following equation:
Equation 2
( BMD ) _hip = 0.78 x ( HU ) _hip + 16.62
In Equation (2), the unit of ( BMD ) _hip is mg / cc.

6. The method of claim 5, wherein the HU value of the lumbar vertebra is obtained from CT (computed tomography) or MRI (magnetic resonance imaging) images without an external phantom.

6. The method of claim 5, wherein the reference bone density for comparison with the bone density measured from Equation (1) is defined by Equation (3)
Equation 3

( BMD ) _phantom is a value measured using quantitative computed tomography ( CTT ) with an external phantom, HU is a value corresponding to an external phantom, the bone mineral density of the lumbar spine from the measurement is the Pearson correlation coefficient (Pearson's correlation coefficient) 0.986 in relation to the (BMD) of equation 3 _phantom.

The method of claim 8, wherein the bone density measured from Equation (1) and the reference bone density of Equation (3) are compared using a root mean square error (RMSE) defined by Equation (4) &Lt; / RTI >
Equation 4

( BMD _phantom ) _i is a BMD value of an i- th voxel calculated using Equation (3), and BMD _phantomless _i is an i- th voxel calculated using Equation (1) (voxel), n represents the total number of voxels, and the RMSE for lumbar spine is 4.26 0.60 [mg / cc].

9. The method of claim 8, wherein said quantitative CT is performed under conditions of 120 kVp, effective mAs 150 mAs, slice thickness of 3 mm and a B40s (medium) kernel. Way.

7. The method of claim 6, wherein the HU value of the femur is obtained from computed tomography (CT) or magnetic resonance imaging (MRI) without an external phantom.

7. The method of claim 6, wherein the reference bone density for comparison with the bone density measured from Equation (2) is defined by Equation (3)
Equation 3

( BMD ) _phantom is a value measured using quantitative computed tomography ( CTT ) with external phantom, HU is a value corresponding to an external phantom, and Equation 2 the femur bone density of the measurement from is the Pearson correlation coefficient (Pearson's correlation coefficient) 0.948 in relation to the (BMD) of equation 3 _phantom.

13. The method of claim 12, wherein the bone density measured from Equation (2) and the reference bone density of Equation (3) are compared using a root mean square error (RMSE) defined by Equation (4) &Lt; / RTI >
Equation 4

( BMD _phantom ) _i is a BMD value of an i- th voxel calculated using Equation (3), and BMD _phantomless _i is an i- th voxel calculated using Equation (2) (voxel), n represents the total number of voxels, and the RMSE for the femur is 8.35 ± 0.57 [mg / cc].

13. The method of claim 12, wherein the quantitative CT is performed under conditions of 120 kVp, effective mAs 150 mAs, slice thickness of 3 mm, and a B40s (medium) kernel. Way.