TP4.5 Precision Without Navigation: Standardized C-Arm Positioning Improves Accuracy in Periacetabular Osteotomy Surgical Model

Abstract

Abstract Introduction Periacetabular osteotomy (PAO) restores coverage of developmental dysplasia of the hip, but is technically demanding, as fragment malpositioning can compromise outcomes. Previous studies demonstrated small changes in pelvic positions significantly alter projected acetabular coverage. While new navigational techniques, such as motion-capture cameras, augmented reality, and patient-specific instrumentation, are emerging, they increase costs and operative complexity. A cost-effective alternative may be intraoperative fluoroscopic alignment matched to preoperative computed tomography (CT) scanning. This study hypothesized that precise C-arm positioning (rotation and flexion/extension) would improve the capability of an experienced surgeon to visually reproduce preoperative CT-based 3D surgical PAO plans. Methods A dysplastic artificial pelvis model mounted with motion-capture tracking was used to simulate PAO and evaluate final fragment orientation. Using custom surgical planning software, 27 CT-based preoperative plans were generated around an “ideal” PAO plan determined in 3D by the surgeon (lateral center-edge angle [LCEA]: 29.3°; Tönnis angle: 0°); 11 were randomly selected for testing. An experienced PAO surgeon visually matched each plan using fluoroscopy under two conditions: (1) preset C-arm alignment, which matched the pelvic tilt and rotation of the 3D preoperative plan using a level, and (2) manual C-arm alignment. Deviations from the plans were quantified using root-mean-square error and standard deviation (RMSE,SD) for rotations, LCEA, Tönnis angle, wall coverage, and estimated range of motion (ROM). Results With matched C-arm positioning, LCEA (3.18°,2.81°), posterior wall coverage (1.17%,1.22%), total coverage (47.54 mm2,42.52 mm2), and ROM (2.15°,1.89°) were not significantly different from planned values (p>0.05). In contrast, manual C-arm alignment resulted in significant deviations from all planned values: LCEA (5.39°,1.80°), Tönnis angle (8.98°,1.45°), anterior wall coverage (5.47%,1.36%), posterior wall coverage (3.75%,0.91%), total coverage (101.39 mm2,38.65 mm2), and ROM (4.32°,1.45°) (p<0.05). Both conditions demonstrated significant rotational error relative to the plan, with inclination (3.80°,2.78° vs 3.07°,1.98°), anteversion (4.49°,1.65° vs 4.04°,1.50°), and flexion (6.36°,2.77° vs 7.95°,2.66°) (p<0.05). Conclusions Precise C-arm position matched to preoperative pelvis alignment improves the accuracy of fluoroscopic guidance in manually reproducing CT-based PAO plans. Although rotational errors persist, standardized alignment more closely achieves planned radiographic parameters, emphasizing the importance of matching pelvis alignment intraoperatively to preoperative plans.