Measurement of Femoral Torsion by Ultrasound and Magnetic Resonance Imaging: Concurrent Validity

Background Abnormal femoral torsion has been linked to osteoarthritis in the knee as well as to patellofemoral pain. Inexpensive, valid, and reliable methods for assessing femoral torsion are needed. Ultrasound (US) is a noninvasive and clinically accessible method that can be used for the assessment of bone morphology, such as femoral torsion. Objective The objective of this study was to determine the concurrent validity of US for the measurement of femoral torsion with a reference method, magnetic resonance imaging (MRI). Design Repeated measurements of femoral torsion were obtained with US and MRI in a laboratory setting. Methods Twenty-eight people (4 men, 24 women; mean age=26.8 years [SD=4.0 years], mean body height=170.3 cm [SD=8.0 cm], mean body weight=64.7 kg [SD=9.8 kg]) participated in this study. T1-weighted axial oblique images of the femoral neck and epicondylar axis were acquired with a 1.5-T magnetic resonance system. Ultrasonographic measurements then were obtained by a tilting technique with a linear transducer that was 4.5 cm long and operated at a frequency of 10 MHz and a depth of 5 cm. Results The average angles of anteversion measured by US and by MRI were 20.7 degrees (SD=11.0) and 19 degrees (SD=11.3), respectively. The reliability, reported as the intraclass correlation coefficient [ICC (2,1)], of repeated measurements of in vivo femoral torsion by US was .98. The reliability [ICC (2,1)] of magnetic resonance image analysis was .96. The standard error of the measurement for US was 2.2 degrees, and that for MRI was 1.9 degrees. The concurrent validity of US with MRI (R2) was .93 (r=.96). Limitations Obtaining measurements by US requires appropriate training before data collection. Conclusions Ultrasound measurement of femoral torsion has high concurrent validity with in vivo MRI and may be used when an assessment of bony morphology is needed but MRI is not available.