3A Keyhole to the Cochlea: Image-Guided and Minimally Invasive Surgery

FThe idea: one straight line to the cochlea

Conventional CI surgery removes a large volume of mastoid bone to open a window onto the middle ear, the facial recess, before reaching the cochlea. Minimally invasive, or percutaneous, surgery replaces that with a single narrow tunnel drilled straight from the skull surface through the facial recess to the cochlea. The trajectory is planned beforehand on a CT scan, choosing a line that threads between the facial nerve and the chorda tympani to land on the round window or scala tympani. The promise is a smaller wound, less bone removal, and a faster, more reproducible approach; the peril is that the planned line passes within a millimetre or less of the facial nerve.[2014][2019]

THow you hit a target you cannot see

In the Vanderbilt percutaneous approach, the planned trajectory is realised with a patient-specific microstereotactic frame: a small custom jig anchored to bone markers that constrains the drill to the single safe line. In the Bern and Antwerp robotic approach, the HEARO robot does the drilling itself, guided by the preoperative plan and the patient's registered anatomy. Registration, matching the live patient to the CT plan, is the make-or-break step; an error of a fraction of a millimetre is the difference between the round window and the facial nerve. Intraoperative imaging, such as cone-beam CT, is taken partway through to confirm the drill is keeping its planned distance from the nerve before it advances to the cochlea.[2014][2019][2022]

TThe facial nerve, and the accuracy that protects it

The facial nerve runs right beside the corridor; injuring it would cause facial paralysis, so the entire technology exists to keep the drill a safe, verified distance from it. In the Antwerp HEARO series the robot achieved mean errors of about 0.18 mm to the target, 0.12 mm to the facial nerve, and 0.11 mm to the chorda tympani, with cone-beam CT confirming safe passage. Even so, the surgeon could complete the robotic procedure in only 22 of 25 planned cases; in the rest, the team converted to a conventional approach, which is exactly the right safety reflex. Continuous facial-nerve monitoring and intraoperative imaging are layered on as backups, because sub-millimetre accuracy still leaves no room for an unverified assumption.[2022][2019]

CWhere it stands today

The Bern first-in-man series showed the keyhole could be robotically drilled to the round window and the patients successfully implanted, proving the concept is feasible in living patients. But this remains largely investigational: case numbers are small, conversions to open surgery happen, and the planning, imaging, and frame or robot add cost and setup time. Not every ear qualifies either; a tortuous facial recess or unusual anatomy may leave no straight line that clears the nerve, ruling the patient out. The honest status is a feasible, regulated, but not-yet-routine approach: powerful where the anatomy is favourable and the team is equipped, and not a replacement for the standard mastoidectomy across the board.[2019][2014][2022]