PURPOSE : The purpose of this preliminary study was to determine the error range compared with preoperative plans in proximal femoral osteotomy conducted using a computed tomography (CT)-based navigation system. METHODS : Four patients (four hips) underwent transtrochanteric rotational osteotomy (TRO), and three patients (four hips) underwent curved varus osteotomy (CVO) using CT-based navigation. Volume registration of pre- and postoperative CT was performed for error assessment. RESULTS : In TRO, the mean osteotomy angle error was [Formula: see text] (range [Formula: see text]) in the valgus direction and [Formula: see text] (range [Formula: see text]) in the retroversion direction. The mean osteotomy position error, with the femoral head side as positive, was -0.4 mm (range -1.4 to 0 mm). The bone fragment rotational movement error was [Formula: see text] (range [Formula: see text]). In CVO, the mean osteotomy position error, with the femoral head side as positive, was -0.2 mm (range -2.0 to 1.7 mm) at the level of the lesser trochanter and 0.8 mm (range 0-3.2 mm) at the level of the greater trochanter. Bone fragment varus accuracy was [Formula: see text] (range [Formula: see text]). CONCLUSIONS : In proximal femoral osteotomy using CT-based navigation, the angle error of osteotomy was within [Formula: see text] and the positional error was within 4 mm. The rotational movement error of the proximal fragment was within [Formula: see text]. These margins of error should be considered in preoperative planning. To improve surgical accuracy, it would be necessary to develop a computer-assisted device which can track the osteotomized fragment.
Keywords: Accuracy; Computed tomography; Navigation; Proximal femoral osteotomy.