Biomedical Engineering Reference
In-Depth Information
field of view when using ionizing radiation, and hence reduce patient dose.
Spectroscopic or perfusion acquisitions can be defined to interrogate spe-
cific tissues of interest, delineated in a previously acquired high resolution
image, rather than a fixed region relative to the scanner. Specific tissue
regions could be followed as the patient moves or is repositioned. These
applications are likely to grow as algorithms become faster and scanner
computing power increases.
Another set of applications relates to use of registration methods to
improve image quality or reduce acquisition time by aligning new data with
previously obtained calibration data or patient-specific information.
16.2.7
Registration of Intraoperative and Preoperative Images in Image-
Guided Interventions
Commercially available image-guided surgery systems are currently
restricted to applications in which patient anatomy can be treated as a rigid-
body, yet this technology has great potential in soft tissues away from bone,
where there is frequently considerable deformation. Several imaging modal-
ities are being developed with interventional applications in mind. These
include MR, CT, x-ray fluoroscopy, and ultrasound as well as optical images
from endoscopes, microscopes, and arrays of free-standing cameras. Regis-
tration methods could be used to update the spatial information in accurate
and detailed representations of the patient generated from preoperative
images using often incomplete and much lower quality information from
intraoperative images.
16.3
Remaining Research Challenges
For all these exciting potential applications of image registration to be real-
ized, several challenges remain, including:
•
Developing a validation methodology for nonrigid registration
algorithms
•
Devising faster and more accurate algorithms (especially non-
rigid). For on-the-fly registration, run times of hundreds of milli-
seconds are desirable
•
Inventing similarity measures that are more robust to image arti-
facts including intensity shading, ghosting, streaking, etc., that can
be applied to more modality combinations
•
Devising algorithms that can distinguish between situations where
tissue volume is preserved or changing, where some structures
are rigid and others are not, and change in one subject over time
compared to differences between subjects
