Stereo-electroencephalography (SEEG) is an invasive procedure, used to identify the epileptogenic zone that can be surgically removed in order to treat drug-resistant epilepsy. Frameless robot-assisted positioning of depth electrodes permits a 3D approach with different obliquities and trajectories. The objective of the present study was to evaluate the morbidity and the accuracy related to this frameless procedure.

Sixty-six patients were managed wherein 901 electrodes were implanted during a 6-year-period. All patients had a postoperative CT-scan that was fused with preoperative MRI planning. In order to assess the accuracy of the procedure, the Euclidian distance was calculated between the coordinates of the planned trajectory and the actual position of the electrode at the entry point and at the target point for 857 electrodes.

Among the 66 patients, one (1.5%) experienced a symptomatic brain haematoma and one (1.5%) a stroke-like migraine after radiation therapy (SMART) syndrome. There was no permanent morbidity or mortality. Compared to the classical SEEG approach, a higher rate of asymptomatic postoperative bleeding was found on the CT-scan in 8 patients (12.1%). Any infectious events were recorded. The median accuracy of frameless robotic SEEG procedure was equivalent to a 1.1mm error deviation (0.15–2.48) at the entry point and 2.09mm (1.06–3.72) at the target point respectively, with no differences for double obliquity trajectories.

Frameless robot-assisted SEEG appears to be a safe procedure, providing sufficient accuracy in order to delineate the epileptogenic zone and represents a helpful tool in the pre-surgical management of refractory epilepsy.