Visually detectable pointer deviation of the nextar surgical navigation system with c-arm, o-arm and ct imaging: a bench-top study using a two-dimensional phantom

Background The accuracy of surgical navigation systems is critical for their safe and effective clinical use. This bench-top study evaluated whether a positional deviation of the pointer tip from a 1-mm target could be detected on the display of the NextAR surgical navigation system (Medacta International SA, Switzerland) when image-to-patient registration was based on C-arm, O-arm or CT imaging, and whether such a detectable deviation depended on the distance between the fiducial reference marker and the target. The endpoint was a visual observation of the pointer on the display and not an independently measured target registration error (TRE). Methods A flat two-dimensional phantom (millimetre graph paper on a wooden board) carried two parallel rows of 20 steel nails with 1-mm heads. Adjacent nails were 10 mm apart and the rows 120 mm apart. The fiducial was fixed midway between the rows at the start of the grid and defined the origin. The phantom was imaged with a Ziehm Vision RFD 3D C-arm, a Medtronic O-arm MVS O2 and a Siemens SOMATOM Drive CT. Ten image acquisitions, each followed by a new image-to-patient registration, were performed per modality. For each nail within the imaged volume, the tip of a navigated pointer was placed on the nail head and the operator visually inspected the NextAR three-dimensional display for a deviation. NextAR provides no numerical deviation value. No artificial registration error was introduced, and the pointer tip was not independently measured. The primary endpoint, visually detectable pointer deviation (hereafter detectable deviation), was a visible offset of the pointer tip from the centre of the nail head on the display. The threshold of about 1 mm is an approximate limit given by the size of the nail head and the millimetre grid and was neither calibrated nor independently measured. Results are descriptive, with exact two-sided Clopper–Pearson 95% confidence intervals. Results No detectable deviation was observed in any acquisition (0/10 per modality; exact 95% CI 0–30.9%; overall 0/30, 95% CI 0–11.6%). The maximum fiducial-to-target distances evaluated were 143 mm with C-arm, 152 mm with O-arm and 199 mm with CT imaging. The intervals refer to acquisitions under fixed bench-top conditions. Conclusions Under the controlled conditions of this two-dimensional bench-top experiment, no deviation was visible on the NextAR display with any of the three imaging modalities. Because the endpoint was a visual assessment limited to about 1 mm and was not independently measured against an external reference, the study does not establish submillimetre accuracy, equivalence between modalities, absence of distance-related effects, or clinical reliability. Verification in anatomically appropriate cadaveric and clinical settings is required. Clinical trial number Not applicable.