EOS low-dose radiography: a reliable and accurate upright assessment of lower-limb lengths.

BACKGROUND Children with lower-limb-length discrepancy require repeated radiographic assessment for monitoring and as a guide for management. The need for accurate assessment of length and alignment is balanced by the need to minimize radiation exposure. We compared the accuracy, reliability, and radiation dose of EOS, a novel low-dose upright biplanar radiographic imaging system, at two different settings, with that of conventional radiographs (teleoroentgenograms) and computed tomography (CT) scanograms, for the assessment of limb length. METHODS A phantom limb in a standardized position was assessed ten times with each of four different imaging modalities (conventional radiographs, CT scanograms, EOS-Slow, EOS-Fast). A radiation dosimeter was placed on the phantom limb, on a portion closest to the radiation source for each modality, in order to measure skin-entrance radiation dose. Standardized measurements of bone lengths were made on each image by consultant orthopaedic surgeons and residents and then were assessed for accuracy and reliability. RESULTS The mean absolute difference from the true length of the femur was significantly lower (most accurate) for the EOS-Slow (2.6 mm; 0.5%) and EOS-Fast (3.6 mm; 0.8%) protocols as compared with CT scanograms (6.3 mm; 1.3%) (p < 0.0001), and conventional radiographs (42.2 mm; 8.8%) (p < 0.0001). There was no significant difference in accuracy between the EOS-Slow and EOS-Fast protocols (p = 0.48). The mean radiation dose was significantly lower for the EOS-Fast protocol (0.68 mrad; 95% confidence interval [CI], 0.60 to 0.75 mrad) compared with the EOS-Slow protocol (13.52 mrad; 95% CI, 13.45 to 13.60 mrad) (p < 0.0001), CT scanograms (3.74 mrad; 95% CI, 3.67 to 3.82 mrad) (p < 0.0001), and conventional radiographs (29.01 mrad; 95% CI, 28.94 to 29.09 mrad) (p < 0.0001). Intraclass correlation coefficients showed excellent (>0.90) agreement for conventional radiographs, the EOS-Slow protocol, and the EOS-Fast protocol. CONCLUSIONS Upright EOS protocols that utilize a faster speed and lower current are more accurate than CT scanograms and conventional radiographs for the assessment of length and also are associated with a significantly lower radiation exposure. In addition, the ability of this technology to obtain images while subjects are standing upright makes this the ideal modality with which to assess limb alignment in the weight-bearing position. This method has the potential to become the new standard for repeated assessment of lower-limb lengths and alignment in growing children. CLINICAL RELEVANCE This study assesses the reliability and accuracy of a diagnostic test used for clinical decision-making.