University of Twente Student Theses


Augmented reality in craniomaxillofacial surgery

Bussink, T.W. (2020) Augmented reality in craniomaxillofacial surgery.

[img] PDF
Abstract:Introduction Craniosynostosis is a congenital disorder that leads to an abnormal cranial shape and possibly increased intracranial pressure. Open cranial vault reconstruction(OCVR) is an surgical method for the treatment of craniosynostosis. The goal of this intervention is to increase the intracranial volume and create a more natural cranial shape. Virtual surgical planning(VSP) provides the possibility to simulate different strategies for the open cranial reconstruction prior to surgery. Although surgical guides can be used to transfer the virtual surgical plan to the patient during surgery, producing these guides is time consuming and rather expensive. Augmented reality(AR) is a 3D visualization technique that can be used to visualize and transfer the virtual planning onto the patient in the operation room in an intuitive, accurate and low-cost way. Materials & Methods To realize the transfer of the VSP, an application for the Microsoft HoloLens, a pair of AR glasses, was developed. Using quick response(QR) codes and a tracking algorithm, point-based registration and realtime tracking were implemented. To accurately transfer the osteotomy pattern, a navigational transfer method was implemented. The created workflow was compared to the conventional method of surgical guides by using 3D prints of 10 patients suffering from different types of craniosynostosis. Each method was tested by two observers at two different moments. To analyze the delineations of the osteotomy patterns on the 3D prints, an improved trilateration method was developed and validated. Results The created application allowed accurate visualization of the osteotomy pattern and reconstruction phase of the VSP. The improved trilateration method showed to be an accurate method for analyzing the delineations. Surgical guides were the most accurate method to transfer the planning to the patient with a mean accuracy of 0.9 ± 0.6 mm and a mean transfer time of 50 seconds. The AR workflow resulted in an accuracy of 2.1 ± 1.5 mm, with outliers up to 8 mm, and a transfer time of 8 minutes and 24 seconds on average. Transfer time and accuracy were significant different between observers. Conclusion Even though the results showed that the AR transfer method is not as accurate as the surgical guides and required a longer transfer time, this method is promising and hopefully clinically applicable for OCVR with the proper training and prevention of outliers. And with the proposed improvements and the introduction of the Hololens 2, AR has the potential to revolutionize the way VSP’s are used and created.
Item Type:Essay (Master)
RadboudUMC, Nijmegen
Faculty:TNW: Science and Technology
Subject:44 medicine
Programme:Technical Medicine MSc (60033)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page