Computer-Assisted Surgery of the Skull Base: Contributions to Guidance and Imaging

E.H.J. Voormolen

Research output: ThesisDoctoral thesis 1 (Research UU / Graduation UU)


Before the 1970s, a skull base tumor implied for the majority of patients a poor prognosis, leading to severe functional impairment and probable death, either from natural progression or surgical efforts. Today, patients treated for lateral skull base tumors have near normal life expectancies in the majority of cases. However, preventing iatrogenic morbidity (e.g. cranial nerve palsy) due to surgery, remains a crucial issue. It is the primary concern of this thesis.

Its chief hypothesis is that application of computer assistance during skull base surgery reduces iatrogenic morbidity. Accordingly, the overall aim of this thesis is to improve the standard of care of patients with lateral skull base tumors by advancing the field of Computer Assisted Surgery of the Skull Base (CASSB).

Specifically, this thesis presents scientific contributions to preoperative imaging and intraoperative image guidance.

In terms of developments in preoperative imaging, it describes the development and evaluation of a novel segmentation algorithm (NerveClick) for semi-automatic segmentation of the intra-temporal facial nerve centerline from high resolution CT images. NerveClick’s is specially developed to be used for image guidance during lateral skull base surgery. Surgeons using NerveClick could segment facial nerve centerlines fast and with high accuracy.

Furthermore, the use of ultra-high field (7T) MR images for image guidance was evaluated. 7T MRI is a promising imaging technology capable of increasing resolution and contrast. For instance, it can be used to improve skull base surgery by enhanced localization of important blood vessels of skull base tumors. We conclude that 7T MR images can be used for CASSB with the following work-around: the patient-to-image registration should be performed on a routine (CT or MR) image, which is subsequently co-registered to the 7T MR image on the image guidance machine.

Concerning advancements in image guidance, this thesis describes methods to offer active real-time intraoperative feedback to the surgeon. It describes the development and validation of new software, called EVADE, which is an acronym for exposure visualization and audible distance emission. It is specifically designed for active image guidance during drilling of the (lateral) skull base. EVADE updates the prior image and visualizes the bone drilling process during skull base surgery virtually in (near) real-time, without need for intra-operative imaging. This functionality is also known as ‘virtual drilling’. Furthermore, the software continuously calculates the distance from the drill tip to segmented normal anatomical structures (e.g. the facial nerve) and produces audiovisual warnings if the surgeon drills in too close vicinity. This feature is called ‘distance control’.

We found that virtual drilling is accurate. Moreover, we conclude that distance control protects normal anatomical structures from iatrogenic damage during drilling of the skull base, at least in an experimental setting. Therefore, we conclude that EVADE has clinical potential. Since we also show that the software is clinically feasible, a randomized clinical trial is needed to confirm the promise of this technology.

Hopefully, this thesis will act as a means to focus research efforts within the field of CASSB and advance it for the benefit of our patients.
Original languageEnglish
Awarding Institution
  • University Medical Center (UMC) Utrecht
  • Robe, Pierre, Primary supervisor
  • Viergever, Max, Supervisor
  • Berkelbach vd Sprenkel, JW, Co-supervisor
  • Noordmans, H.J., Co-supervisor
Award date10 Jul 2018
Print ISBNs978-94-6233-983-5
Publication statusPublished - 10 Jul 2018


  • neurosurgery
  • image guidance
  • skull base
  • 7 Tesla magnetic resonance imaging


Dive into the research topics of 'Computer-Assisted Surgery of the Skull Base: Contributions to Guidance and Imaging'. Together they form a unique fingerprint.

Cite this