FROM CT IMAGE TO 3D MODEL | ||||
by Y. Ted Wu, Ph.D.
(Published on Advanced Imaging, August 2001, pp. 20-23) | ||||
Thanks to new 3D imaging software, surgeons can
now create physical models of their patients' inside. | ||||
Mr. A. Waitz broke the radius of his right forearm in an automotive accident five years ago. It was poorly set initially and has been reset several times. He has been to some of the top surgeons in the US and had nine restorative surgeries so far. Because of the lack of good radiological information, the physicians were not able to achieve good alignment and hence the failure of all these surgeries. Multiple surgeries would result higher risk to the patient, as well as more costs and significant additional recovery times. In the case of nine surgeries with a typical minimum recovery time (length of time in a cast) of 6 weeks, it requires 54 weeks or more than a full year of time in recovery, as well as starting physical therapy again each time. Surgery today
is still similar to what it did three thousand years ago: the surgeons use
their hands to directly control instruments and they use their eyes to
provide them with feedback about the effect of their manipulations. The
amount of information they have before a surgery is critical to the
success of the operation. The developments in 3D imaging technology over the recent years have been focusing on the goals to not only let the physicians and surgeons to see better inside the body, but also create physical models from CT (Computed Tomography) and MRI (Magnetic Resonance Imaging). Using the model a surgeon is able to perform a "mock" surgery prior to ever entering the operating room. Some surgeons find that minutes or hours of operating time can be saved by careful preparation using the model. | ||||
Rendering a Surface Model | ||||
Able Software Corp. 3D-DOCTOR (Lexington, MA) is a 3D imaging software using vector-based surface model creation rather than raster-based systems to significantly improve computational efficiency and the quality of generated 3D models to be used by 3D printers and rapid prototyping machines. | ||||
![]() |
Click on Picture to Enlarge* | |||
First, the image is segmented to isolate and trace
the boundaries of the object of interest within the image set (Figure 1).
A single CT or MRI image slice (512x512 16 bit) is about a half a megabyte
in size and a typical CT/MRI image can have around 200 slices, which makes
the total image size about 100 megabytes. Figure 2 shows a CT pelvis image
of a patient. Rather than work with all the pixel values as conventional
3D modeling systems do, 3D-DOCTOR defines the boundaries of objects in
vectors -- a closed polygon with a given magnitude for all image slices.
To create the 3D polygon based surface model, the surface rendering
algorithm searches the best corresponding point in the neighbor slices to
connect for forming surface triangles. When all boundary points are
properly connected, a watertight surface model is created. Thanks to the
vector-based surface rendering algorithm, a 3D surface model can be
created in a few seconds on a standard Pentium III class PC with 128 MB
memory. The 3D volume rendering can be done in near real-time (from a few
seconds to few minutes depending on the complexity of the object under
study) without using any additional hardware accelerator. By comparison,
other raster-based medical imaging software packages recommend
workstations with dual P3’s and up to 1 GB of RAM, some of them have to
use special volume rendering hardware for 3D volume display. The 3D
surface model can be visualized on a PC in real time from any angle. The
model can also be sliced or diced to show the image voxels on the model or
at certain position within the body. | ||||
Creating The Physical Model | ||||
The 3D models are then saved to a file format that a 3D printer or a rapid prototyping machine accepts. For example, the STL (stereolithography) file format is commonly used on most 3D printers or rapid prototyping machines. Other standard formats include AutoCAD DXF, IGES and 3D Studio file formats. | ||||
![]() |
Click on Picture to Enlarge* | |||
A Rapid Prototype
Machine takes the
surface model and automatically creates a physical model using material
such as resin, paper, powder or plastics. This process can take minutes to
hours depending on the complexity of the model. Most 3D printers are still
monochrome, using the same color as the material used. However, colors
start to appear on some of the newer 3D printers and pretty soon multiple
colors will be available for producing complex models. Once the model
is made, the surgeon not only can look it from any angle, but also touch
it and even operate on it. Using the model a surgeon is able to perform a
"mock" surgery prior to ever entering the operating room. Some surgeons
find that minutes or hours of operating time can be saved by careful
preparation using the model. The model is a great tool for a surgeon to communicate with her patient about the upcoming procedure. Instead of looking at a textbook drawing or a generic anatomical model, the patient gets to visualize the procedure on their own model. An added benefit is that the model becomes an ideal communication tool between surgeons and specialties on the case at hand. Unlike film or drawings, the model can only be interpreted basically one way, significantly reducing confusion or misunderstanding in the communication process. | ||||
Real-Life Applications | ||||
Doctors at the
Massachusetts General Hospital Surgical Simulation Group (Boston, MA) are
using 3D-DOCTOR to make replicas of real internal organs from CT images.
Under a US Army grant, the group is combining a 3D surgical visualization
with a life size mannequin of a human torso to create a “flight simulator
for surgeons.” The system will be used to train US Army medical personnel
on the best ways to handle penetrating battle wounds, such as insertion of
tubing into the lungs to alleviate air pressure from gun wounds.
As for Mr. A. Waitz,
he started his own searching for imaging tools to help his doctors better
prepare and plan a surgery. He has found 3D-DOCTOR and is currently
working with his doctors to use the software to create 3D models for both
his left arm (normal) and the right arm (injured) to prepare another
surgery. Hopefully with the help of the 3D imaging technologies and the
real models, this will be the last surgery he’ll ever need--well, for this
particular injury, at least. Y. Ted
Wu, Ph.D., is the founder, President and CEO of Able Software Corp.
in Lexington, MA. He is also the architect and developer of 3D-DOCTOR
software, which is a FDA-approved for medical imaging
applications. |