Laparoscopic Soft Tissue
Scanning/Registration
Spatially registered 3D
preoperative medical images can
improve surgical accuracy and reduce reliance on memory and
hand-eye coordination by the surgeon. They enable visualization
of internal structures within the anatomy of a patient on
the operating table. In the case of biopsy, for example,
this would allow the surgeon to guide the needle tip to a tumor
though opaque tissue. It has been well established that for
soft tissues, image registration can be performed if the
surface contours of the organ can be sensed during surgery. That is,
one can align
the preoperative images (e.g. CT, MRI) with a cloud of
points that
describe the surface of an organ. The surgeon can then use this
information during the surgery to see exactly where his tool (knife,
scissors, needle, etc.) is with respect to internal structures beneath
the surface of the organ, BEFORE beginning to cut tissue.
Prior systems for soft tissue registration have required wide exposure
(aka "open surgery") of internal organs for surface sensing. Our
contribution is a scanning system that can accurately collect a cloud
of points describing organ surface contour minimally invasively -
through a laparoscopic port. To do this, we combine a 1D laser distance
measurement device, the Optimet Conoprobe,
with a Micron optical tracker. This system enables a non-contact 3D
surface scan
to be collected through a laparoscopic port.
(Left) The conceptual sketch of
conoscope surface scanning system. (Right) Experimental apparatus
used in initial proof-of-concept studies.
The images below show the results of a registration experiment
using an anthropomorphic liver phantom. A CT scan
of the phantom provided the green and orange surface. The
blue point cloud was taken from a conoscope scan of the same
phantom. An Iterative Closest Point (ICP) registration was then
used to
register the point cloud to the CT generated surface. The phantom
and the registration results are shown below.
Conference Publications:
- Lathrop RA, Webster, R J
III,
2009, “Conoscopic Holography
for Image Registration: A Feasibility
Study", Proceedings of SPIE, 2009.