Target registration and target positioning errors in computer-assisted neurosurgery: proposal for a standardized reporting of error assessment.

BACKGROUND: Assessment of errors is essential in development, testing and clinical application of computer-assisted neurosurgery. Our aim was to provide a comprehensive overview of the different methods to assess target registration error (TRE) and target positioning error (TPE) and to develop a proposal for a standardized reporting of error assessment.

METHODS: A PubMed research on phantom, cadaver or clinical studies on TRE and TPE has been performed. Reporting standards have been defined according to (a) study design and evaluation methods and (b) specifications of the navigation technology.

RESULTS: The proposed standardized reporting includes (a) study design (controlled, non-controlled), study type (non-anthropomorphic phantom, anthropomorphic phantom, cadaver, patient), target design, error type and subtypes, space of TPE measurement, statistics, and (b) image modality, scan parameters, tracking technology, registration procedure and targeting technique.

CONCLUSIONS: Adoption of the proposed standardized reporting may help in the understanding and comparability of different accuracy reports.