Related Experiment Videos
Three-dimensional image registration of phantom vertebrae for image-guided surgery: a preliminary study
Diane M Muratore1, Jeannette Herring Russ, Benoît M Dawant
1Department of Biomedical Engineering, Western New England College, Springfield, Massachusetts 01119, USA. dmurator@wnec.edu
Summary
This study demonstrates that free-hand, tracked three-dimensional ultrasound (3D US) images can be accurately registered to computed tomography (CT) scans for spinal image-guided surgery (IGS). This technique allows for precise surgical targeting using minimal vertebral surface data.
Area of Science:
- Medical imaging
- Surgical navigation
- Biomedical engineering
Background:
- Three-dimensional ultrasound (3D US) applications are expanding in medicine.
- Accurate registration of 3D US to other modalities like CT is crucial for image-guided surgery (IGS).
- Precise pre-operative targeting can improve surgical outcomes in spinal procedures.
Purpose of the Study:
- To develop and evaluate a method for registering tracked, free-hand 3D US images of a phantom spine to computed tomograms (CT).
- To assess the accuracy of this registration technique for potential application in spinal IGS.
- To provide an independent, quantitative measure of registration accuracy.
Main Methods:
- Acquired 3D free-hand US images of a phantom spine using optical tracking.
- Segmented the posterior vertebral surface from both US and CT images.
- Performed surface-based registration between the segmented US and CT data.
- Quantified registration accuracy using a fiducial marker system.
Main Results:
- Feasible alignment of 3D US and CT image sets using only portions of the vertebral surface (spinous process and laminae).
- Target registration errors (TREs) were lower with higher resolution CT scans (2-mm slice thickness).
- TREs were 1.33 ± 0.30 mm (7.5-MHz US) and 2.81 ± 0.10 mm (4.5-MHz US) using 2-mm CT slices.
Conclusions:
- Rigid, surface-based registration of CT to tracked 3D US spinal phantom images is achievable.
- Effective registration can be accomplished using limited, easily accessible portions of the vertebral surface.
- This method shows promise for enhancing precision in spinal image-guided surgery.