Dose, image quality and spine modeling assessment of biplanar EOS micro-dose radiographs for the follow-up of in-brace adolescent idiopathic scoliosis patients

Purpose The aim of this study was to compare the radiation dose, image quality and 3D spine parameter measurements of EOS low-dose and micro-dose protocols for in-brace adolescent idiopathic scoliosis (AIS) patients. Methods We prospectively included 25 consecutive patients (20 females, 5 males) followed for AIS and undergoing brace treatment. The mean age was 12 years (SD 2 years, range 8–15 years). For each patient, in-brace biplanar EOS radiographs were acquired in a standing position using both the conventional low-dose and micro-dose protocols. Dose area product (DAP) was systematically recorded. Diagnostic image quality was qualitatively assessed by two radiologists for visibility of anatomical structures. The reliability of 3D spine modeling between two operators was quantitatively evaluated for the most clinically relevant 3D radiological parameters using intraclass correlation coefficient (ICC). Results The mean DAP for the posteroanterior and lateral acquisitions was 300 ± 134 and 433 ± 181 mGy cm 2 for the low-dose radiographs, and 41 ± 19 and 81 ± 39 mGy cm 2 for micro-dose radiographs. Image quality was lower with the micro-dose protocol. The agreement was “good” to “very good” for all measured clinical parameters when comparing the low-dose and micro-dose protocols (ICC > 0.73). Conclusion The micro-dose protocol substantially reduced the delivered dose (by a factor of 5–7 compared to the low-dose protocol) in braced children with AIS. Although image quality was reduced, the micro-dose protocol proved to be adapted to radiological follow-up, with adequate image quality and reliable clinical measurements. Graphical abstract These slides can be retrieved under Electronic Supplementary Material.