The problem of mismatches between CT scan and DXA results

Abstract

Background. It is reported that radiodensity measured in Hounsfield units becomes more and more popular in bone property assessment, however also mismatch with DXA results is observed.Purpose. The aim of this study is to evaluate the relationships between the results of DXA and CT with a focus on explanations for observed discrepancies.Material and Methods. This is a cross-sectional study; forty patients were enrolled, all patients underwent DXA and CT. A bone mineral density BMD (g/cm2) was calculated for each vertebra of a lumbar spine (L1-L4 inclusive), neck, upper neck, shaft, Wards triangle and trochanter of hip. Bone radiodensity in HU was taken from each vertebral body in the sagittal, axial and coronal planes. A total vertebra body radiodensity including cortical bone and radiodensity of only cancellous bone were calculated. To assess a potential impact on DXA and CT data agreement a mean radiodensity and square of the right and left vertebral pedicles and facet joints were measured for each vertebra.Results. A strong correlation between BMD measured using DXA and CT data was estimated with a multiply r accounting for 0.84169, p < 0.0001, however the most contributing parameters were those calculated for facet joints. It has been detected that both radiodensity of only a cancellous bone and total have a weak correlation with matching BMD measurements of a proximal femur.Conclusion. The results of DXA could be strongly influenced by hypertrophic changes of facet joints. Both CT and DXA measurements taken from a lumbar spine may have a mismatch with figures taken from hip.

Keywords

• computed tomography
• osteodensitometry
• radiodensity
• bone mineral density