Phantom-less volumetric vertebral density obtained from routine abdominal CT studies: correlation with data obtained by DXA

Abstract

Conventional quantitative computed tomography (QCT) uses a calibration phantom scanned simultaneously with the anatomical region of interest and measures bone density accurately and with short-term high precision. Evidence supports that phantomless volumetric BMD highly correlates with QCT BMD and is a reliable method for assessing bone density of vertebral bodies. Assessment of BMD in routine abdominal CT scans has been investigated in recent years. The aim of the study was to correlate BMD and bone mineral content (BMC) obtained from CT studies with data obtained by DXA. Twenty eight women (age 63.4± 10, 3 years old, range: 37-85) who underwent abdominal CT for different reasons and DXA measurements within 6 months were included. A simple manual region of interest (RI) which delineated the edge of the vertebral body was applied to L3.We measured 1) CT: Volumetric integral density (BMDv) -trabecular and cortical bone- of the axial section in Hounsfield units (UH) and area (A) in cm2. Density was multiplied by area to obtain a value equivalent to BMC. 2) DXA: BMD and BMC in a RI of 10 mm height in the middle of L3 All parameters obtained by CT correlated significantly with the corresponding to DXA : BMDv vs BMDa r: 0.67 (p=0.005) y BMC- CT vs BMC-DXA: r :0.75 (p=0.00063). This study complements previous reports and opens the possibility of using routine abdominal CT studies to assess bone density. For that purpose reference values (age and gender) must be established.