Image-guided surgery: what is the accuracy?

Abstract

Use of image-guided surgery (IGS) systems in otolaryngology, particularly rhinology, has grown exponentially in recent years. Central to their use is the understanding of the accuracy of each system. The purpose of this review is to discuss the error inherent in all IGS systems. A standardized technique (currently used in the engineering literature) for understanding and reporting error in IGS systems is reviewed. Using this technique, the error of commercially available IGS systems is reviewed. The most commonly used IGS systems depend on the conformation of the skin, as opposed to relying on bone-implanted devices. For these systems, mean accuracies 2 mm or less are routinely reported. This finding is independent of fiducial markers (eg, proprietary headsets, skin-affixed markers, or laser scanning of skin surfaces). Techniques of fiducial localization and registration of CT scans to intraoperative anatomy are proprietary to each company. As such, there is great variability in reporting system specifications-particularly error of IGS systems. This lack of standardization makes comparison of one system to another difficult if not impossible. Image-guided surgery systems commonly used in rhinology report mean accuracies of 2 mm or less. Surgeons must be aware that this value represents a mean of a distribution of errors. As such, 95% of the time error can be expected to be less than approximately 1.7 times its mean value. However, outliers (errors much larger and much smaller than the mean) may exist for each IGS intervention. As noted, IGS systems function to complement-not replace-knowledge of surgical anatomy.