Computer-assisted fracture reduction: novel method for analysis of accuracy.

Anatomic reduction of displaced fractures is limited by the chosen surgical approach and intraoperative visualization. Preoperative Computed Tomography (CT) enhances the analysis of the fracture pattern and provides accurate spatial relationships. Computer Assisted Surgery (CAS) was introduced to increase the accuracy of specific surgical procedures. CAS systems can be used for implant placement or osteotomies in intact bone or reduced situations prior to obtaining the CT data, as differentiation into different datasets related to specific fragments is not yet possible. We present a model that allows "virtual" controlled reduction, providing computer assistance during the fracture reduction. Prior to clinical application, the accuracy of the process of virtual reduction must be proven in an experimental setting. An in vitro fracture model with two body fragments and a motion tracking system for three-dimensional (3D) control (accuracy 0.1 mm and 0.1 degrees ) was used. Two methods were employed: direct visualization and reduction by the examiner, and "virtual" reduction, performed solely with the use of a computer image, in which the examiner lacks any direct visualization of the fragments. The results of this very simplified "fracture" model indicate that the overall difference between direct and virtual controlled reduction was very small. A significant difference of 0.3 mm (0-1.8 mm) was seen for the residual displacement represented by the Euclidean distance (p < 0.01), whereas the difference in the residual angulation was not significant (p > 0.05). The methods tested revealed that virtual controlled reduction is nearly as accurate as direct visualization. Reduction control utilizing a motion tracker system reveals accurate 3D information in this simplified reduction setup, and is now used as a standard setup for analyzing realistic fracture models. Copyright 2001 Wiley-Liss, Inc.