Introduction: Malrotations following Several complications have been reported in femoral nailing, among them. The aim of this study is to develop an intraoperative method based on cone beam CT (CBCT) to assess comminuted fracture periaxial rotation. We hypothesize that bone surface matching using CBCT image data can precisely predict malrotation in the fractured femur even with severe comminution.

Methods: A mid-shaft osteotomy in a fresh frozen cadaveric femur was performed and a rotational axis was formed. The proximal part of the femur was fixed and the distal part was optically racked for periaxial rotation. At each rotation a CBCT was aquired. The images were segmented at bone threshold. The center of the bone in each axial slice was calculated and the distance from that center to the inner and outer bone surfaces was sampled at 1o intervals (360x). The resulting plot was an unwrapped virtual bone surface consisting of a pattern of ridges and valleys. Fracture gaps were simulated by removing CT slices adjacent to the osteotomy. The fracture gap was reconstituted using an extrapolation algorithm to the midline of the fracture. The two bone surfaces were then continuously shifted relative to one another in order to match the geometric bony features. Calculated malalignments were compared to the measured at each of the 16 rotations with each of the 9 simulated fracture gaps. Three rotational malrotations were tested twice to assess repeatability.

Results: Femoral malrotation was strongly predicted as compared to the rotation measured by optical tracking. The performance was not impacted by gap size up to 100 mm.

Discussion: The high quality of intraoperative CBCT imaging data enables surface matching algorithms to be utilized. The results ratify this novel method for assessing fracture rotation.