A framework for determining component and overall accuracy for computer assisted surgery systems

摘要

Computer assisted surgery (CAS) systems have generated a great deal of interest in the research and medical communities due to their promise of improved results with less invasive access to the surgical site. While some systems have been checked against a minimum acceptable standard set by the system's developers, no overall methodology is in place to insure that a system is thoroughly characterized. A framework to assess the accuracy of CAS systems will provide developers and researchers with a common starting point, and will provide surgeons with a consistent basis for comparing different systems. This framework is described, and the application of the framework is demonstrated with the HipNav system. Background: Multiple groups have explicitly modeled outcome errors for their CAS systems, looking at the final surgical or test result in comparison to the ideal result. Other groups have examined individual subsystems for their component accuracies. Simon described accuracy validation for image guided orthopaedic surgery [2]. He identified sources of error for prototypical CT-based systems, and suggested techniques to minimize the effects of some of the sources. Viant described a graph-based method to model the error propagation through a CAS system [3].