Feasibility of a customizable training environment for neurointerventional skills assessment

Abstract

Objective: To meet increasing demands to train neuroendovascular techniques, we developed a dedicated simulator applying individualized three-dimensional intracranial aneurysm models ('HANNES'; Hamburg Anatomic Neurointerventional Endovascular Simulator). We hypothesized that HANNES provides a realistic and reproducible training environment to practice coil embolization and to exemplify disparities between neurointerventionalists, thus objectively benchmarking operators at different levels of experience.

Methods: Six physicians with different degrees of neurointerventional procedural experience were recruited into a standardized training protocol comprising catheterization of two internal carotid artery (ICA) aneurysms and one basilar tip aneurysm, followed by introduction of one framing coil into each aneurysm and finally complete coil embolization of one determined ICA aneurysm. The level of difficulty increased with every aneurysm. Fluoroscopy was recorded and assessed for procedural characteristics and adverse events.

Results: Physicians were divided into inexperienced and experienced operators, depending on their experience with microcatheter handling. Mean overall catheterization times increased with difficulty of the aneurysm model. Inexperienced operators showed longer catheterization times (median; IQR: 47; 30-84s) than experienced operators (21; 13-58s, p = 0.011) and became significantly faster during the course of the attempts (rho = -0.493, p = 0.009) than the experienced physicians (rho = -0.318, p = 0.106). Number of dangerous maneuvers throughout all attempts was significantly higher for inexperienced operators (median; IQR: 1.0; 0.0-1.5) as compared to experienced operators (0.0; 0.0-1.0, p = 0.014).

Conclusion: HANNES represents a modular neurointerventional training environment for practicing aneurysm coil embolization in vitro. Objective procedural metrics correlate with operator experience, suggesting that the system could be useful for assessing operator proficiency.

Fig 1. Image of HANNES in the research laboratory.

The simulator is fabricated of different units including the electric and control element, the fluid system including a water tank, the standardized aorta and the patient-specific cervical and intracranial vessels.

Fig 2. Bar charts illustrate catheterization times, operator experience, aneurysm models and attempts.

(A) Experienced operators showed a clear trend towards shorter catheterization times than inexperienced physicians. Both groups needed longer to catheterize the most challenging aneurysm model #3. (B) Mean catheterization times per attempt for all aneurysm models, grouped by the operator’s experience. Faster catheterization times were observed for both inexperienced and experienced operators with consecutive attempts, but this correlation was significant for the inexperienced physicians only.

Fig 3. Spectrum of results and complications during coil embolization.

(A) Final framing coil position in aneurysm model #1, coil placement performed by an inexperienced operator, result graded as insufficient (3). (B) Erratic wire movement against the aneurysm wall of model #2 during aneurysm catheterization by one inexperienced operator. (C) Dislocation of the microcatheter from the aneurysm sac during placement of the framing coil, performed by an inexperienced operator. (D) Dislocation of the framing coil into the parent vessel during coil placement, performed by one inexperienced operator. (E) Coiling result of one inexperienced operator, showing a residual neck (Raymond-Roy Occlusion Classification II), arrow indicating coil prolapse into the ICA. (F) Coiling result of the most experienced operator, showing a residual neck (Raymond-Roy Occlusion Classification II).