The integration of robotics and computer control have changed manufacturing drastically. These technologies may be poised to do the same for medicine. Researchers are designing robots for operating-room tasks with the goal of providing better and more cost-effective care than can be provided now by unaided surgeons. Already, a few companies have developed commercial systems to aid surgeons.
The promise of using computers and robots is not merely about providing new tools. Russell Taylor, director of the Center for Computer-Integrated Surgical Systems and Technology at Johns Hopkins Univ., points out that developers need to consider entire systems rather than a single robot tool. Taylor writes, "The introduction of robots into surgery will change surgical processes just as profoundly as robots changed manufacturing processes."1
Why use robots at all?
Figure 1. ZEUS (left) is a robot designed for minimally invasive surgery; a surgeon controls tools from a console (right), rather than holding them directly. This has been used for a variety of surgeries, including a heart bypass surgery in which the chest wall was not opened and the patient's heart kept beating.
Humans have excellent judgment, adaptability, and dexterity. And because humans communicate well, a surgeon can easily tell a human helper what to do. Humans also have excellent hand/eye coordination. Despite the fact that human surgeons are capable of remarkable feats, however, there are some areas in which a machine can perform better.
Robots don't get tired, inattentive, or succumb to coffee jitters; they can be designed to work at very small scales and without the inherent muscle tremors that limit human steadiness; they can be designed to work in spaces too small for human hands, and perform tasks that require geometric accuracy very well; and robots can be sterilized and are immune to ionizing radiation, which can be useful for surgeries that use x rays taken during the procedure.
Ultimately, of course, surgical robots will be designed to cooperate with surgeons, rather than replace them. A surgeon in the future might use computer programs that manipulate pre-operative images of the patient and help to plan a procedure.2 (There is even an acronym for this: CASP, or computer-aided surgical planning.) Once the surgery begins, the doctor may use robots with integrated sensors to provide real-time feedback and registration of the pre-operative images. Finally, the surgeon can use robotic actuators in procedures that exploit their strengths. For example:
- minimally invasive surgery where human hands and eyes won't fit;
- surgery on a scale too small to be comfortable for unaided human hands;
- procedures such as milling hip bones that require more geometric accuracy than humans can reliably provide.
A major roadblock to the use of robots in operating rooms is capital cost, and the perception that adding technology to the operating room will drive up costs. Taylor disagrees. He writes, "Advances in technology can reduce both the direct costs of treatment and the indirect costs to society of untreated disease and deformity by (1) reducing the morbidity and hospitalization associated with invasive procedures; (2) reducing surgical complications and errors; (3) improving consistency; and (4) making possible cure or amelioration of hitherto untreatable conditions at an affordable price." Challenges
Developers of robots to aid surgeons face different design criteria than developers of industrial robots. The safety requirements for medical devices are far more stringent than for standard industrial machines, and robots are no exception. Both rigorous testing and redundant safeguards are necessary for these applications. Designers need to consider two possibly dangerous failure modes: component failure and the possibility of a robot executing a correct command at the wrong place or time.
In particular, the actuators must be designed with safety in mind. Taylor writes, "Great care needs to be taken to protect both the patient and operating-room personnel from run-away conditions." He believes that modifications from standard industrial robot designs should include redundant position sensing, lowering the maximum speed of movement, and evaluating -- and possibly redesigning -- the robot for electrical safety and sterility.
Sterility is a major issue for medical robots. Most of the robot can usually be draped or covered with a sterile bag, leaving only the instruments or end-actuators to be sterilized. Sterilization by gas or soaking tend to be less destructive to equipment, but also less common, than using an autoclave.
For minimally invasive procedures, the actuator sizes must be designed to pass through a small opening in the patient's body. Commercial robots for minimally invasive surgery
Given these requirements for surgical robots, one might be surprised at how far these devices have already been developed. In addition to research robots in development at a number of universities and hospitals, several companies offer commercial devices.
In July the FDA announced a change in the way robotic devices are cleared for market release. Robotic devices that perform surgical tasks such as cutting and suturing will be cleared for market release under the FDA 510(k) Premarket Notification process (which has a statutory review time of 90 days) instead of the Premarket Approval process (which is a 180-day process that requires FDA Advisory Panel presentations).
This announcement is good news for companies such as Computer Motion in Santa Barbara, CA, which makes the ZEUS Robotic Surgical System for minimally invasive microsurgery procedures (Figure 1). In September 1999, Canadian surgeon Douglas Boyd used ZEUS to perform heart bypass surgery endoscopically, without stopping the patient's heart. ZEUS is not cleared for marketing in the U.S. for bypass surgery, however.
In July, the FDA cleared another device, the Da Vinci Surgical System, for marketing. The Da Vinci, made by Intuitive Surgical, Inc. (Mountain View, CA), enables a surgeon to perform laparoscopic gall bladder and reflux disease surgery while seated at a console with a computer and video monitor. The surgeon uses hand grips and foot pedals on the console to control three robotic arms that perform the surgery using a variety of surgical tools.
Figure 2. The robot ROBODOC mills bone to precisely fit a prosthesis in hip replacement surgery, which improves the bonding of the bone to the prosthesis. This reduces problems with dislocation as well as reducing patient pain and recovery time.
Integrated Surgical Systems (Sacramento, CA) created a robot system called ROBODOC to aid surgeons conducting hip replacements by performing computer-controlled shaping of the patient's femur to the prosthesis (Figure 2). A good fit is important because ideally the bone will grow into the porous end of the implant and fix it in place -- but the bone will only grow about 0.5 mm. The hand tools used by surgeons in conventional approaches are difficult to use precisely. According to ROBODOC's manufacturer, "Conventional surgical techniques routinely leave gaps of 1 mm or more between the bone and the implant."
The cutting edge used in ROBODOC is a high-speed cutting burr, computer controlled to cut a pattern determined before surgery using ISS software and a computer tomography scan of the patient.
The system has not been cleared for marketing in the U.S., but has been approved in several European countries. In Germany earlier this year, the ROBODOC system was successfully used for knee replacement surgery.
Many more robots are likely to follow. Their development will be inextricably bound with the development of computer planning and imaging tools, as well as other applications, such as telemedicine.
1. R. H. Taylor, "Medical Robotics and Computer-Integrated Surgery" in the Wiley Handbook of Robotics, 1999.
2. Much more information about computer-aided surgery is available on the Web. The Computer Aided Surgical Systems and Technology site at http://cisstweb.cs.jhu.edu is a good place to start.
Yvonne Carts-Powell, based in Boston, writes about optoelectronics and the Internet.