gms | German Medical Science

60th Annual Meeting of the German Society of Neurosurgery (DGNC)
Joint Meeting with the Benelux countries and Bulgaria

German Society of Neurosurgery (DGNC)

24 - 27 May 2009, Münster

Comparison of freehand three-dimensional ultrasound reconstruction techniques for image guidance in neurosurgery

Meeting Abstract

  • D. Miller - Klinik für Neurochirurgie, Universitätsklinikum Essen
  • C. Lippert - Fachhochschule Gießen-Friedberg
  • F. Vollmer - BrainLAB
  • S. Hertel - Institut für Medizinische Informatik, Universitätsklinikum Essen
  • D. Schult - Klinik für Neurochirurgie, Universitätsklinikum Essen
  • U. Sure - Klinik für Neurochirurgie, Universitätsklinikum Essen

Deutsche Gesellschaft für Neurochirurgie. 60. Jahrestagung der Deutschen Gesellschaft für Neurochirurgie (DGNC), Joint Meeting mit den Benelux-Ländern und Bulgarien. Münster, 24.-27.05.2009. Düsseldorf: German Medical Science GMS Publishing House; 2009. DocP02-11

doi: 10.3205/09dgnc271, urn:nbn:de:0183-09dgnc2718

Published: May 20, 2009

© 2009 Miller et al.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License ( You are free: to Share – to copy, distribute and transmit the work, provided the original author and source are credited.



Objective: Freehand three-dimensional ultrasound imaging (3D-US) is increasingly used in image-guided surgery. During image acquisition, a set of B-scans is acquired that is distributed in a non-parallel manner over the area of interest. Reconstructing these images into a regular array allows 3D-visualization. The reconstruction step is very important as any loss of image quality has to be avoided. The aim of the study is to compare different algorithms with respect to image guidance in neurosurgery.

Methods: 3D-US data sets were acquired during surgery of various intracerebral lesions using an integrated ultrasound-navigation device (BrainLAB). They were stored for post-hoc evaluation. Five different reconstruction algorithms, a standard multiplanar reconstruction (MPR), a pixel nearest neighbor method (PNN), two different voxel nearest neighbor methods (VNN and VNN2) and a distance weighted algorithm (DW1) were tested.

To test the robustness of the algorithms to fill gaps within the sample volume, we removed various amounts of input data and checked the ability of the algorithms to recreate the removed data in a grey value analysis in four different patient data sets. Secondly, a total of 450 reconstructed images from five patients was evaluated with respect to image quality by two different neurosurgeons that were blinded towards the reconstruction algorithm. Grades from 1 (very good) to 6 (very bad) were given depending on the ability to define tumor borders, ventricles and dural structures in each image. The two best algorithms were compared to the original 2D ultrasound images from a further eight patient data sets in the US plane of view to check for artifact production within the image.

Results: Grey value analysis showed that MPR was significantly worse than the other algorithms in filling gaps when more than 1% of the input data was removed (p<0.05). When evaluating the diagnostic value of reconstructed axial, sagittal and coronal images, VNN2 and DW1 were judged to be significantly better than MPR and VNN. Mean grades (standard errors) were 2.0 (0.12), 2.0 (0.13), 2.8 (0.24), 3.5 (0.25), respectively and 2.4 (0.18) for PNN. No relevant artifact production could be seen in VNN2 and DW1 in comparison to the original 2D images.

Conclusions: VNN2 and DW1 could be identified as robust algorithms that generate reconstructed US images with a high diagnostic value. These algorithms improve the utility and reliability of 3D-US imaging during intraoperative navigation.