Augmented Reality for Fluoroscopy-based Navigation
A generalized framework for surgical implants and instruments into
our Open Inventor based modules has been built. Software was developed
for the conversion of CAD drawings and inclusion of 3D volumetric
datasets into our framework. Based on this framework an AO/ASIF
foundation grant was funded for a database with all major Synthes®
trauma implants and instruments. The framework was connected to this
database to achieve full 3D virtual planning and surgery.
On-the-fly computation and visualization of X-ray free fluoroscopic
image updates was another focus of this subproject. A prototype C-arm
navigation system was designed to incorporate X-ray free image updates
and the interactive use of a Synthes® trauma implant, the
LISS® plate, a long bone fracture reduction, and a fixation
device system. With the help of the virtual cylinders from different
X-rays, a bone structure of similar shape and volume was created. The
reduction process can be checked without continuous radiation.
Moreover, the implants, including the plates, screws and
instruments are compared three-dimensionally to virtual X-rays,
realistically adjusted and represented in real-time. The insertion of
the plate and the fixation with screws, including the drilling process
and linear measurement, are precisely navigated in a 3D virtual
world. This system was successfully tested in the laboratory
setting. The system was then further enhanced to semi-automatically
generate the virtual cylinder scene based on gradient vector field
based segmentation of the long bones. Several tibial long fracture
cases have been successfully treated with our system.
Also, research in C-arm based precision placement of acetabular
components during total hip replacement (THR) was performed (this work
was awarded the HAP Paul award from the International Society of
Technology in Arthroplasty). The THR system was clinically evaluated
and an accuracy of better than 5 degrees inclination and 6 degrees
anteversion was achieved under clinical conditions, which implies that
there is no significant difference in performance from established
CT-based navigation methods. The work has been adapted by our partner
Medivision/Praxim® into a clinical module that is in routine use.
|