Volume II, No. 1
Winter 1991
  
ROBODOC... The Development Of Robotics in Medicine
 Like most new ideas, the use of robotics was born of frustration. Most of you know of  our interest in custom designed cementless hip replacements. We have developed a method which we believe is extremely accurate in designing and manufacturing the femoral component (the part that goes into the thigh bone). You might think that since these prosthesis are custom designed to each patient they should be easy to implant-"just slip  right in." In fact,the opposite is true. The implants are designed to fit very tightly to the hard bony wall of the femur. Cutting and shaping this bone to match the implant exactly is a difficult task-more difficult, in fact, than other Off-the-shelf" implants that are not intended to fit so precisely.

The tool used to shape the upper femur for nearly all implants, whether off-the-shelf or custom, is called a broach or rasp. It is a tool that is the exact size and shape of the implant, but has cutting teeth on it. After the femoral head and neck have been removed, the surgeon uses a heavy mallet to drive this cutting tool into the upper end of the femur after the femoral head and neck have been removed.

In the case of a custom prosthesis, a custom broach is made for each patient. Since a custom prosthesis is designed to fit tighter to the hard bony wall, the process of broaching the femur for a custom prosthesis is more difficult and time consuming.

So we were frustrated! We had designed something we felt was more accurate and would fit better, but the process of preparing the femur for implantation was more difficult. We also suspected that because of the large forces generated by the mallet and the fact that the broach was handheld, there might be some inaccuracies created in the process of preparing the bone. Subsequent studies have shown this to be the case.

The idea then came to me that if we were using three-dimensional analysis to design and manufacture the prostheses using CAD/CAM (Computer Assisted Design/Computer Assisted Manufacture) techniques, we might be able to use the process to prepare the bone for implantation. The computer uses an NC (numerical controlled) machine to cut out the prosthesis from a solid block of Titanium alloy. What we needed was a computer-controlled device that could use the same information and cut a matching surface inside the femur. Unfortunately, NC machines typically are quite large, weigh over a ton, and are very dirty. Not exactly what you'd want in an operating room.

A robot arm is also computer controlled, is much smaller, and can be made to function in an ultra-clean environment. We thought we could put a high speed cutting burr on the end of a robot arm and use the reverse of the cutting tool paths that the NC machine uses to  make the prostheses.

I discussed the idea with Dr. Hap Paul, my veterinarian research partner. He agreed that it seemed feasible to use a robot arm to do the job in surgery. What we didn't know was that robots are basically stupid They follow relatively simple commands and are mainly used in industry to do a single task over and over and over.

It turned out that no one had developed a way to easily program a robot to do a new and very complicated task that each time was different (like cutting a shape to match a custom prosthesis). we approached most of the large robot manufacturing firms-either they couldn't do it or were not interested. Again, frustration.

I was discussing my problems one day with my father-an old IBMer who'd spent 40 years  with that company. He said "I can't believe IBM doesn't have someone somewhere working on programming robots." He called some old friends at IBM and soon I got a call from the head of the Robotics Research Section at the IBM Thomas Watson Research Center in Yorktown Heights, NY. He was a little excited because he'd gotten a call from the Chairman of the Board of IBM who asked him to call me. It turned out that IBM had developed a new computer language specifically designed to program robots to do new and complicated tasks.

In 1985, IBM hired Brent Mittelstadt, my graduate student. Brent spent a year in Yorktown on IBM's payroll doing a feasibility study. He came to the conclusion that it was definitely feasible.

Over the next four years, Dr. Hap Paul led the team through the complicated application and development process. This was done at UC Davis under an IBM grant of money and equipment.

Now the research and development is over, and it's time to take this technique into the operating room. On May 5, 1990, Dr. Paul and I did the first robotic total hip replacement on a dog that was crippled with arthritis. That was quite a day-culminating five years and many man-hours of work.

The process as it now is performed consists of several steps. It begins with a minor surgical procedure, which can be done as an outpatient. Three small titanium pins (smaller than a dime) are implanted (by the surgeon, not the robot) into the femur bone; one on each side. Just above the knee and a third near the hip. Then a CT scan is performed. The CT information is used to design the implant if a custom prosthesis is to be used.

The CT information is then fed into a computerized imaging system we developed with IBM called "ORTHODOCK". This system finds the pins and allows the surgeon to decide where he desires to place the implant in the bone. If an off-the-shelf implant is to be used, the surgeon can use ORTHODOCK to three-dimensionally decide what size implant to use and where it would best fit in the bone. Once the surgeon has manipulated the implant into the desired position on the computer screen, he pushes a button and ORTHODOCK automatically registers the implant's position relative to the three pins.

At the time of surgery, the surgeon performs the first part of the procedure normally.  When he is ready to prepare the femur for the femoral component, the leg is placed in a special holder or fixator which holds the femur rigidly. The robot is then brought in and connected to the fixator. The three pins are exposed and the robot is guided to the pins.

Using the information from ORTHODOCK, the robot finds the exact center of each pin. It then knows where the surgeon wants the implant placed. The robot, using a high speed burr, cuts out the desired shape inside the femur working from the top down. The surgeon then implants the femoral component and the procedure is finished routinely. The use of the robot adds about 15 to 30 minutes to the procedure.

Many studies of accuracy and safety have been performed. The accuracy is felt to be 10  to 40 times better than the handheld manual broaching technique described earlier. Multiple  layers of redundancy are used for safety and error recovery. Ultimately, however the surgeon has a big red button that can stop the process if something occurs that he doesn't like. The surgeon watches what is occurring on a real-time graphics monitor. The surgeon is still in control and makes all the judgments both before and during the procedure.

Is the surgeon's job in jeopardy? I don't think so. This is merely a new tool to perform a part of the procedure more accurately than handheld broaches. The surgeon will always be needed for Judgment and control.

What about the future? We hope to perform the first case on a human in mid- 1991 at Sutter General Hospital. The FDA and other regulatory hurdles still remain to be crossed. Meanwhile, we are mapping out research into other areas, such as placement of the acetabular component, total knee replacement, osteotomy of bones, ligament reconstruction in sports medicine, etc.

Although we have developed the idea of using a robot in surgery to aid in performing hip replacements, we have been careful to keep it generic so that it could be adapted easily to other types of surgery. We have plans to explore its application to ophthalmology, ENT, neurosurgery and cancer surgery. There is a potential application of this technology for any  procedure that can be planned three-dimensionally using a CT or MRI scan and where improved accuracy of placement and cutting are required.

Finally, it is our hope that new procedures before not thought possible, will become possible as a result of the development of this new technology. Never, in our wildest dreams did we ever envision this when we first began.

Return to top

Office Staff Profiles
We would like to introduce a key person in the administration of our office, manager, Anna Adair. Unfortunately, her office is located away from the reception and examination rooms so patients don't always get to see her when they arrive for appointments.

Anna is a native Californian, born and raised in the high desert town of Lancaster. After  working in the public relations department of Home Savings in Los Angeles, she relocated to the small town of Oakdale, in Stanislaus County. Commuting to Modesto, she worked for National Medical Enterprises (NME) and transferred to Sacramento while still with NME. Looking for a change, she accepted the position of office manager with Dr. Bargar. Her work
with him began in May of 1986 to prepare his private practice to open in July. The position encompasses a variety of duties: overseeing office personnel, coordination of Dr. Bargar's travel, speaking and meeting engagements, and basic accounting practices associated with a business.

When not at work, Anna enjoys time with her nine year old daughter, Rachael, camping, movie watching, as well as snow and water skiing.

Return to top