Computer-assisted Surgery With Bone Anchoring in Otologic Surgery.

OBJECTIVE

Computer-assisted navigation (CAN) is increasingly used in ENT surgery, notably in sinus and otoneurosurgical procedures. This study assessed the accuracy of a CAN system in temporal bone surgery using bone-anchored registration, without preoperative fiducial screws or facial CT imaging, focusing solely on the temporal bone.

MATERIALS AND

METHODS

Ten cadaveric temporal bones (5 left, 5 right) underwent high-resolution CT imaging. Dissections (mastoidectomy with facial recess approach) were conducted using the Collin Navigation Solutions system. Surface registration employed the iterative closest point (ICP) algorithm based on 3 anatomic landmarks. Positional accuracy (PA) was evaluated for 3 middle ear structures: short process of the incus (INC), head of the malleus (MAL), and anterior margin of the round window (COC) and one internal ear structure: lateral semicircular canal (LSC). PA was defined as the distance between each anatomic point and its corresponding CT location.

RESULTS

Forty measurements were obtained. Initial PA was <1 mm in 77.5% of cases: 70% for both LSC and MAL, 90% for INC, and 80% for COC. The mean PA was 0.67 ± 0.46 mm before reregistration, and improved to 0.44 ± 0.32 mm after reregistration on the LSC. This improvement was statistically significant for LSC ( P = 0.0048) and INC ( P = 0.046), but not for MAL ( P = 0.2443) or COC ( P = 0.952).

CONCLUSION

This study showed encouraging results regarding the accuracy of the CAN system for temporal bone surgery and supports further investigation.