Fiducial-based registration with a touchable region model
- 298 Downloads
Image-guided surgery requires registration between an image coordinate system and an intraoperative coordinate system that is typically referenced to a tracking device. In fiducial-based registration methods, this is achieved by localizing points (fiducials) in each coordinate system. Often, both localizations are performed manually, first by picking a fiducial point in the image and then by using a hand-held tracked pointer to physically touch the corresponding fiducial on the patient. These manual procedures introduce localization error that is user-dependent and can significantly decrease registration accuracy. Thus, there is a need for a registration method that is tolerant of imprecise fiducial localization in the preoperative and intraoperative phases.
We propose the iterative closest touchable point (ICTP) registration framework, which uses model-based localization and a touchable region model. This method consists of three stages: (1) fiducial marker localization in image space, using a fiducial marker model, (2) initial registration with paired-point registration, and (3) fine registration based on the iterative closest point method.
We perform phantom experiments with a fiducial marker design that is commonly used in neurosurgery. The results demonstrate that ICTP can provide accuracy improvements compared to the standard paired-point registration method that is widely used for surgical navigation and surgical robot systems, especially in cases where the surgeon introduces large localization errors.
The results demonstrate that the proposed method can reduce the effect of the surgeon’s localization performance on the accuracy of registration, thereby producing more consistent and less user-dependent registration outcomes.
KeywordsRegistration Fiducial marker Surgical navigation Surgical robot
This work was supported by NSF NRI 1208540.
Compliance with ethical standards
Conflict of interest
The authors declare that they have no conflict of interest.
A human subject study was performed, with institutional review board approval (HIRB00003967), to evaluate the proposed registration method. No animal studies were performed.
This article does not contain patient data.
- 8.Maurer CR Jr, Fitzpatrick JM, Wang MY, Galloway RL Jr, Maciunas RJ, Allen GS (1997) Registration of head volume images using implantable fiducial markers. IEEE Trans Med Imaging 16(4):447–462Google Scholar
- 10.Nagy DA, Haidegger T, Yaniv Z (2014) A framework for semi-automatic fiducial localization in volumetric images. Augment Environ Comput Assist Interv 8678:138–148Google Scholar
- 12.Pieper SD, Halle M, Kikinis R (2004) 3D Slicer. In: IEEE International Symposium on Biomedical Imaging (ISBI), 15–18 April, Arlington, VA, USA. pp 632–635Google Scholar