by Jean
Thilmany, Associate Editor
|
These days, liver surgeon Rory McCloy's operations
are guided by a four-foot, three-dimensional virtual profile of a patient's
liver that rotates at his command on the operating room wall in front
of him. Using a specially outfitted mouse and his laptop computer, McCloy
can manipulate the virtual liver to find the exact location and extent
of a tumor before he makes even one incision in the patient's actual liver
on the table below.
This virtual medical technology isn't ahead of its time, McCloy said.
Instead, it has been slow to be adopted by the medical establishment and
he, for one, aims to speed the pace. McCloy is a surgeon at the Manchester
Royal Infirmary at the University of Manchester in England.
In the past 20 years, virtual reality and virtual imaging have revolutionized
the way physicians practice medicine, McCloy said. Magnetic resonance
imaging, computer-aided tomography scans, and ultrasound images give doctors
a vital peek into the inner workings of a patient's body in a way that
X-rays cannot. Not only can these imaging tools diagnose disease, tumors,
and the like, but they serve as a map to each person's unique body. Just
as people are different on the outsidethough they all have a nose,
two eyes, and two earsinternal organs are also unique. Being able
to see exactly what an organ looks like before operating allows doctors
to perform closer and more precise work.
Liver
surgeon Rory McCloy of the University of Manchester Royal Infirmary in
England adapted virtual image technology to study three-dimensional images
of the patient's liver as he operates.
McCloy is charged with cutting tumors from the liver and the pancreas.
Nicking the tumor during surgery can release cancerous cells into the
organ, so he needs to map his cuts.
But MRI, CT, and ultrasound techniques only touch on what technology is
capable of now, McCloy said. The shame of it is that more virtual reality
technology isn't exploited for the medical environment, he said.
As a doctor long interested in virtual reality, McCloy has gone to meetings
and conferences on that subject for the past eight years and has seen
firsthand the feasibility of using 3-D scans for more than just diagnostic
purposes. The scans could be used in the operating arena to take the place
of the X-rays hung around the room to help guide his knife.
"Here I am at a state-of-the-art teaching hospital in the U.K., one
of the biggest hospitals in Europe, and I'm still given X-ray films or
CT scans that are slices, like slices of salami," he said. "I
have to reconstruct in my mind, in the operating room, the 3-D image."
In fact, many hospitals in the United States digitally reconstruct a patient's
body or organ in 3-D by use of a CT scan across the body, he said. But
those reconstructions are often available only at a computer workstation
in the X-ray room.
"The surgeons go down there and look at it and then vanish to the
operating theater, and still have to remember where the tumor is,"
he said. "And while they're very good at remembering, it's the difference
between cutting here or cutting three millimeters away in hitting cancer
or not hitting cancer."
McCloy likens his operations to extracting an entire plum from a pie with
only a few knife slices allowed. It would be easier to do that if you
were guided by images in three dimensions that exactly mirrored the pie,
rather than looking at slices and trying to figure out how big the plum
was and exactly where it was situated.
"I'm busy in my mind during operations trying to put the X-rays into
3-D and going into the patient's body and trying to figure out where this
tumor is, and it's all a bit demanding," McCloy said. "Since
we're in the digital age, wouldn't it be nice to use all the digits?"
Before using the virtual reality technology, which he and Nigel John,
a computer scientist at the university, prepared for the operating room,
McCloy studied patients' CT scans and X-rays before surgery to decide
how to best operate. During surgery, he didn't have access to the CT scan,
because it needed to reside on a computer and the images often didn't
load fast enough to really help a doctor.
"On the X-ray, I can see the tumor on one slide and then on another
slide, and I can work out that it's several millimeters long," he
said. "The X-ray doctors are happy to look at the slide and see that
there's a tumor there at that particular part of the liver. That's the
diagnosis. But I'm faced with six pounds of liver or the pancreas, which
is the size of the banana and buried deep in the body, and you can't see
it because it's hidden behind a lot of other structures.
"I know the tumor is there because the X-ray doctors told me it was
there," McCloy continued. "But it's a bit like a lucky dip.
You put your hand into a barrel and pick out the sweeties. With the X-ray,
you know there are six packages of sweeties in the barrel, but you don't
know exactly where they are. Only it's not a barrel, it's a liver that
bleeds when you cut it. And I have to remove the cancer whole because
if I cut into it, the cells spread around the body."
Why Can't I See a Liver in 3-D?
Because of his affiliation with John and because he has attended conferences
on virtual reality, McCloy knew that the technology he sought was available.
However, he had to find his own way to make it available to him, to get
it to meet his own needs. John helped with that.
"My frustration was sheer frustration that people kept giving me
these X-ray films while the data has been there for years just sitting
on the scanners," he said. "It's okay to look at a lovely 3-D
image, but why can't I have that while I'm operating? I'm the guy that
has to get the 3-D tumor out."
Because the amount of data contained in a series of CT scans is too large
for the graphics card on a personal computer to handle, McCloy couldn't
simply bring a computer into the operating room and refer to the patient's
CT while operating. The scan would take forever to load. So he and John
found a way to send the CT information to a supercomputer at the University
of Manchester computing center more than a mile from the hospital. The
supercomputer was then networked with a laptop McCloy used in his operating
room.
The team used an Onyx 300 visualization computer from SGI, formerly Silicon
Graphics Inc., of Mountain View, Calif. The machine is specially configured
for high-powered graphics imaging by use of the company's InfiniteReality
graphics capability. It simultaneously processes two-dimensional images,
3-D graphicsin this case, the CT scanand data in real time.
A team at the on-campus Manchester Visualisation Centre wrote the software
that allows the raw 50 to 80 megabytes of data from the CT images to be
graphically visualized.
The
image of the patient's liver that McCloy follows while operating is composed
of the patient's CT scan, which is sent to a university supercomputer
and then to McCloy's laptop computer.
The high-powered graphics are available to McCloy by remote access on
his decidedly non-supercomputer, thanks to software called OpenGL Vizserver,
also from SGI, which allows the data from the supercomputer to be shifted
to other computers, even much less powerful ones, via a network link.
In this case, the laptop in McCloy's operating room serves as a remote
station that lets him see and manipulate the CT in 3-D while operating.
Because the images take no time to load, he can manipulate them in real
time. For instance, he can rotate the image of the liver.
To keep the files small for negligible loading time, only the pixels of
the graphics are transmitted to the laptop, not the graphics themselves.
The computer generates about 20 images a second. McCloy uses a common
projector, the same one he uses for PowerPoint presentations, to display
the image on the wall.
Because the graphic is what's called a volume reconstruction of the entire
data setin this case, an image of the entire liver, both inside
and outMcCloy can slice the imagethe liverfor a 2-D
look inside the organ to mirror the X-rays he formerly saw. He likened
the technology to a 3-D anatomy book customized for each patient and projected
at a height of six feet on the operating-room wall. The wall is, in McCloy's
words, a tasteful shade of pale green and the images show up clearly even
under the bright operating-room lights.
Practicing Before Operating
McCloy performed his first liver surgery using the technology in April,
a second in May, and a third in July. The next surgery will be done on
a pancreas rather than on a liver. And he's been using the technology
to plan surgeries before going into the operating room.
"But the exciting bit is getting the scan on the wall while operating,"
he said.
Because
images take no time to load on his laptop, McCloy can manipulate them
in real time. They're reconstructions of the 3-D liver, so he can slice
them to see inside the organ.
Still, McCloy runs through the surgery on another laptop computer before
going into the operating theater.
"I can practice before I make the cut," he said. "I can
say to myself, ÔIf I cut it at this angle, will it work?' And I
can see on-screen that it won't. Then, I can cut through it another way
and see that, yes, I can get it out this way."
Still, the scan won't accurately reflect the shape of the real liver McCloy
will see because organs aren't rigidly conformed inside the body. They
move around a bit and spread, or are contracted by other organs and their
shape changes accordingly, he said.
"When I actually have the patient's organ in front of me, it may
not be in the same shape or at the same angle it is on screen," he
added. "But I can rotate the organ on screen so it shows up at the
same angle."
Joystick Leads the Change
The trick in making the technology useful for the operating room lay in
finding a way to manipulate the images with one hand while performing
the surgery with the other, said John, of the school's computing
center. A traditional mouse is slow to work with because of the sanitary
requirements of the operating room. The mouse can't be sealed within
a sterile plastic bag to meet operating room requirements because the
roller ball doesn't work so well when encased in plastic, John
found.
McCloy has to put on a sterile glove before using a standard mouse in
the operating room. When he needs both hands to operate, he takes off
that glove and puts on another. It's a time-consuming process that
he expects a plastic-encased joystick to eliminate.
A joystick works fine encased in plastic. John's team is now linking
an $18 joystick, the type used for computer gaming, to the software program.
Many university teaching hospitals have been working with SGI to find
uses for visualization technology, according to Chodi McReynolds, director
of marketing at the SGI sciences division. One area of continued study,
for example, is pairing a CT scan with customized graphics of the patient's
blood flow through veins and arteries provided courtesy of computational
fluid dynamics software, McReynolds said.
The pairing of a patient's body scan with graphics that depict
blood flow could be used to study how surgery would affect that particular
person.
"If they wanted to do a bypass in one spot, they could do that
on the image, and it would show how blood flow to the leg would be affected,"
she said.
For their part, McCloy and John hope to network their supercomputer to
other laptops in operating rooms throughout the United Kingdom. First,
McCloy will use the technology during 30 surgeries over the next year
to document its usefulness. In the next few weeks, a gynecologist colleague
will begin looking at ways a digitized and projected CT scan of the pelvis,
ovaries, and uterus might be used during surgery.
The visual technology that McCloy and John implemented was funded by a
$25 million grant from the Department of Education in the U.K.
"This is a collaboration between a computer scientist, a surgeon,
and a company that has the power of computation with large amounts of
number crunching," McCloy said. "Put the three of us together,
shake us up, and we've come up with this as an answer."
home |
features |
news update |
marketplace |
departments |
about ME |
back issues |
ASME |
site search
© 2002 by The American Society of Mechanical Engineers
|