gms | German Medical Science

69. Jahrestagung der Deutschen Gesellschaft für Neurochirurgie (DGNC)
Joint Meeting mit der Mexikanischen und Kolumbianischen Gesellschaft für Neurochirurgie

Deutsche Gesellschaft für Neurochirurgie (DGNC) e. V.

03.06. - 06.06.2018, Münster

Millimetric augmented reality intraoperative registration using signature vessels: a technical report

Meeting Abstract

  • Lima Sprenger - Hopitaux Universitaire de Genève, Neurosurgery, Genf, Schweiz
  • Ivan Cabrilo - Schulthess Klinik Zürich, Spine Surgery, Zürich, Schweiz
  • Karl Schaller - Hopitaux Universitaire de Genève, Neurosurgery, Genf, Schweiz
  • Philippe Bijlenga - Hopitaux Universitaire de Genève, Neurosurgery, Genf, Schweiz

Deutsche Gesellschaft für Neurochirurgie. 69. Jahrestagung der Deutschen Gesellschaft für Neurochirurgie (DGNC), Joint Meeting mit der Mexikanischen und Kolumbianischen Gesellschaft für Neurochirurgie. Münster, 03.-06.06.2018. Düsseldorf: German Medical Science GMS Publishing House; 2018. DocV283

doi: 10.3205/18dgnc301, urn:nbn:de:0183-18dgnc3013

Veröffentlicht: 18. Juni 2018

© 2018 Sprenger et al.
Dieser Artikel ist ein Open-Access-Artikel und steht unter den Lizenzbedingungen der Creative Commons Attribution 4.0 License (Namensnennung). Lizenz-Angaben siehe http://creativecommons.org/licenses/by/4.0/.


Gliederung

Text

Objective: Augmented reality (AR) is modifying the perception of a real-world environment by adding real-time computer-generated information. Applied to neurosurgery, valuable information is implemented in the surgeon's operating field, forgoing the need to regularly turn to the navigation system. Current technologies suffer drawbacks regarding ergonomics, complexity, precision and trust. Here we focus on improving the visual experience, the main issue being its lack of sobriety, leading to distraction.

Methods: 103 interventions were performed using signature vessels' millimetric registration (aneurysm clipping N=87, AVM resection N=5, extra-intra cranial bypass for revascularisation N=2, glioma resection N=2, cavernoma resection N=2, dural fistula exclusion N=4, foreign body extraction N=1). Skin surface and skull were segmented as reference objects preoperatively. Characteristic vessel segments possibly visible in the operative field were segmented as signature vessels. Pathological structures were segmented as target objects. Co-registered patient and microscope reference arrays allowed AR image injection into the microscope's eyepiece. After registration and accuracy verification, a custom-made mini craniotomy was performed. Vessels seen while progressing were used to verify the perfect spatial correlation of the virtual structures with reality, allowing micro-readjustments.

Results: Objects searched were unambiguously identified and mismatch recognized. Readjustments were easy and accurate to the limitation of brain pulsations. The ability to identify and verify in real time the relative position of modeled and visible structures increased the surgeon's trust in the extrapolation of structures not readily seen in the operative field. The rendering of virtual images allowed the surgeon to operate without the need for removing the overlay even at high magnification.

Conclusion: Precision is limited by the indirect nature of the setup relying on fiducial tracking. Furthermore, reliability is limited by the elastic nature of brain tissue and disruption of rigid co-registration. Millimetric registration of signature vessels coupled with frequent intraoperative micro-adjustments allow to improve these drawbacks.