Long bone fracture reduction using a fluoroscopy-based navigation system: A feasibility and accuracy study

Abstract

Objective: Long bone fracture reduction during intramedullary nailing can result in a small but significant rate of angular and rotational malalignment, which can in turn lead to long-term morbidity. Current techniques for intraoperative reduction rely heavily on fluoroscopy, and their reproducibility can be limited. We suggest that fluoroscopy-based navigation may improve the precision of long bone fracture reduction and reduce radiation exposure.Methods: A cadaveric tibia was stripped of soft tissues and fractured at its midline. The ends were then cemented to two mobile brackets within a fracture/deformity simulator. Optical trackers were drilled into each fragment. Radiographs were obtained including AP and lateral views of the proximal and distal ends of the bone as well as the fracture site. These radiographs were stored in the computer navigation system. Fracture reduction was performed using a fluoroscopy-based navigation system with virtual intraoperative planning software. The system used 4 sets of lines drawn by the surgeon on the fluoroscopic AP and lateral images. While navigating the reduction of the fracture these lines aligned together, providing graphic and numerical descriptions of the reduction. The lines included the anatomic axis of the bone, the matching of the fracture lines, the short segment anatomic axis (near the fracture site), and the AP mid-sagittal joint line (MSJL) of both the plateau and the plafond. Anatomic reduction was then performed and the computer assessment of the angulation and translation of the fragments was recorded. Each stage was repeated 25 times for each set of lines. For the control group the surface of the bone was tracked using an optical probe.Results: The accuracy of the system varied according to the planning method. The most accurate technique was matching of the fracture lines, which yielded 2.68 ± 1.18 mm of translation and 2.5 ± 1.27° of angulation. The next most accurate method was using the short segment anatomic axis, followed by using the anatomic axis method. The control group was the most accurate (1.78 ± 0.67 mm translation, 1.58 ± 0.97° angulation). The AP MSJL yielded errors greater than 10°.Conclusions: Fluoroscopy-based navigation is sufficiently accurate for long bone fracture reduction, potentially increasing angular and translational accuracy while reducing the amount of intraoperative fluoroscopy. Navigation may improve the outcome of treatment of a long bone fracture by better restoring the normal mechanical axis of the limb in a less-invasive closed manner.