Accuracy of peripheral quantitative computed tomography and magnetic resonance imaging in assessing cortical bone cross-sectional area: a cadaver study.

OBJECTIVE: The present study aimed to assess the accuracy of peripheral
quantitative computed tomography (pQCT) and magnetic resonance imaging (MRI) in
assessing the cortical cross-sectional area (CCSA) at the level of the tibia.
METHODS: Nine human whole-leg specimens were scanned with pQCT and MRI
(T1-weighted and ultrashort echo-time [UTE] images) at 2 diaphyseal levels, 1
distal and the other 1 proximal. Subsequently, the bones were cut, and the CCSA
was measured using digitized planimetry on the transverse sections. The
correlation between CCSAs assessed via pQCT and MRI (T1 and UTE images) and
assessed via planimetry, taken as a criterion standard, was evaluated using the
Spearman rank correlation method.
RESULTS: The mean (SD) CCSA with pQCT was 237.3 (54.3) mm2; T1-MRI, 228.8 (63.2)
mm2; UTE-MRI, 178.5 (54.8) mm2; and planimetry, 250.17 (59.3) mm2. The CCSA
measured with pQCT and T1-MRI was highly correlated with the planimetric CCSA
(rho = 0.868 and rho = 0.880, respectively, P <0.001). The correlation involving
UTE-MRI was somewhat weaker (rho = 0.664, P = 0.003).
CONCLUSIONS: Peripheral quantitative computed tomography and MRI (T1-weighted
images) are accurate for the assessment of the CCSA at the tibial shaft.