3D Modeling of Lower Extremities With Biplanar Radiographs: Reliability of Measures on Subsequent Examinations.

BACKGROUND: Biplanar radiography with 3-dimensional (3D) modeling (EOS) provides a comprehensive assessment of lower limb alignment in an upright weight-bearing position with less radiation than conventional radiography. A study was performed to assess the consistency and reliability of 2 lower extremity 3D biplanar radiograph models created at least 1 year apart in a pediatric population.

METHODS: All patients who had 2 lower extremity radiographic evaluations with EOS performed at visits a minimum of 1 year apart were reviewed. Digital radiographs, of lower extremities in both frontal and sagittal planes, were acquired simultaneously, using the EOS system. The 3D reconstruction of the images was achieved utilizing the SterEOS software. Pelvic position, femoral and tibial anatomy, and the torsional profile were evaluated and compared using t tests.

RESULTS: In total, 53 patients with a mean age of 11.7 years (range, 6.1 to 18.9 y) met inclusion criteria. When comparing 3D models between visits, minimal differences were noted in proximal femoral anatomy and pelvic alignment (pelvic incidence, sacral slope, sagittal tilt, neck shaft angle). Expected differences in femoral and tibial length corresponded with normal longitudinal growth between visits. Sagittal plane knee position varied widely between examinations. Femoral and/or tibial rotational osteotomies were performed in 37% of extremities between examinations. After femoral derotational osteotomy, a significant difference in femoral anteversion was appreciated when comparing preoperative and postoperative 3D models. However, this difference was less than the expected difference based on the anatomic correction achieved intraoperatively. No differences were noted in tibial torsion measures after tibial derotational osteotomy.

CONCLUSIONS: The 3D modeling based on biplanar radiographs provides consistent and reliable measures of pelvic and hip joint anatomy of the lower extremity. Patient positioning may influence the reproducibility of knee alignment. The torsional profile assessment did not accurately reflect changes obtained by derotational osteotomy.

LEVEL OF EVIDENCE: Level III.