Prosthetic Knees: What’s on the Way?
October 2003 Issue
It has been said that there's nothing new under the sun, but somehow there always seems to be a new modification, approach, or philosophy that, when applied to an "old" concept, certainly appears to create something new. And why worry about semantics when the result is something that works better, faster, or easier to improve the lives of lower-limb amputees?
What's waiting to debut now--and what could be coming in the more distant future?
Otto Bock's C-Leg has already undergone numerous changes since its debut six years ago, said Greg Schneider, CP, clinical specialist in prosthetics, Otto Bock Health Care, Minneapolis, Minnesota. "Some changes have been fairly major, such as adding a second mode into the C-Leg for use in sports and other kinds of activities; others have been somewhat invisible tweaks to make things work better and more efficiently. It will continue to evolve in a similar fashion--we'll just keep making it a better and better product."
What's in R&D for the C-Leg can't be discussed, Schneider said, but microprocessors continue to be an improving technology in O&P. "There will be lots of exciting things coming up; prospects are virtually limitless as batteries keep getting better," he exclaimed. "I can envision every prosthesis having a microprocessor someday--but that's just my personal vision." Also in the design stages at Otto Bock: new approaches to the 20-year-old weight-activated stance mechanism for the safety knee. With the old weight-activated stance brakes, the user had to take the weight off to get the knee to flex for the swing phase. Some of the new developments will have weight-activated stance mechanisms that don't need to be unweighted to flex the knee during stop for pre-swing and swing phase. "Those types of knees will be out soon," said Schneider.
Looking toward the future, Fillauer Inc., Chattanooga, Tennessee, is examining other ways of providing stance control in knees, such as the application of other kinematic, mechanical, or position-dependent locks, said Gerry Stark, BSME, CP, FAAOP, director of education and technical support. "Besides locks which represent better involuntary controls, more should be done to enhance voluntary control." He hinted that new products can be expected in these areas within a year or so.
Ohio Willow Wood, Mount Sterling, Ohio, is also cautious about revealing too much, but Mark Ford, director of marketing, noted that the company "will continue to develop more new versions of the GeoLite and GeoFlex knees based on the same technology. Such versions should be introduced within the next 12 months."
Regarding its new knee, Ossur North America, with offices in Aliso Viejo, California, stated the company did not want to comment at this time and hopes to keep the knee under wraps until its introduction. Also in the wings on Ossur's stage is a motorized prosthetic knee soon to be ready for market. The active powered knee is in final development stages at Victhom Human Bionics in Quebec, Canada, Ossur's new partner. Ossur will act as manufacturer and distributor of Victhom's prosthetic products, and will co-develop subsequent O&P products using Victhom's cutting edge bionic technology.
Unlike passive lower limb components, Victhom's knee is capable of active flexion and extension, which should make it more energy-efficient and more natural in motion than currently available knees. A computer will control the powered knee's movement, selecting from a range of normal walking parameters which were decoded and computerized by Victhom's founder and chief scientific officer, Stephane Bedard. Victhom reportedly anticipates its commercial availability within a few months.
Endolite North America, Centerville, Ohio, was also wary about revealing new product directions. "At this moment, it's early days," said Alan Kercher, technical service and education manager. "I don't really want to comment on some things that are in progress. Blatchford is a very innovative company that likes to push the boundaries; we're always looking at new possibilities," Kercher hinted.
Sandia National Laboratories, Albuquerque, New Mexico, is still hard at work on its Smart Integrated Lower Limb. Its microprocessor will not only control the knee, but will manage, process, and coordinate input from the ankle, foot, and socket. Reportedly, technical requirements for the limb will be set by Seattle Systems, Poulsbo, Washington, while materials work and testing will be performed by Sandia's peacetime partners in the Russian nuclear weapons laboratory, Chelyabinsk 70.
The leg is intended to simulate a human gait regardless of irregular or steep uphill/downhill terrain. A microprocessor-controlled module implanted in the leg will respond to sensor input from multiple sources; the microprocessor will control hydraulic joints and electric motors that power the ankle, knee, and socket. The leg is thus "smart" enough to behave responsively, reacting to current conditions.
"This is about making a leg that is more like a missing limb than a collection of components ever can be," said a Sandia spokesperson. "This limb will have a digital control system to make it smart."'
Also coming soon: A very unusual prosthesis, perhaps best labeled a "false prosthetic," was described by Jack Engsberg, PhD, director of the Human Performance Laboratory, Department of Neurological Surgery, Washington University, St. Louis, Missouri:
"Our objective is to create a training aid that will help teach physicians and clinicians to assess strength and spasticity. This will take the form of a mannequin fitted with artificial joints at the knee, elbow, ankle, and hip. The joints can be set for different levels that mimic strength and spasticity levels encountered in human patients."
Patient spasticity and strength can vary according to time of day, anxiety level, motivation, fatigue, age, gender, size, and strength. Current methods for determining spasticity require manually moving a patient's passive limb through a range of motion and assessing the resistive characteristics (spasticity) based upon six grade descriptions that are largely subjective. Methods for determining strength rely on the patient isometrically resisting applied loads; the evaluator senses, then grades the resistive isometric force of the patient. Again, the determination is somewhat vague. "So where first responders can practice their skills on a CPR dummy, our students will practice assessing spasticity and strength correctly on our training mannequin--which can also be used for testing and certification of their assessment skills," Engsberg explained.
Lord Corporation, Cary, North Carolina, is working with Engsberg to develop an appropriate resistive brake for the mannequin's artificial knee joint. The prosthetic knee is the first joint to be developed using magnetorheological (MR) fluids that can instantly change viscosity characteristics to provide specific resistive torque at a fixed angle or a torque profile over an entire angular range of motion. Engsberg anticipates that the spasticity dummy will be available in about three years.
The Future: What's Possible?
In what direction do you anticipate prosthetic knee design to evolve in the future?
Gerry Stark, Fillauer: "We are all looking at different ways of triggering the stance control mechanism. In the past, most of the triggers have been load-related. "Right now, what people are looking at are different linkages, braking systems, positional control, and electronic triggers. I believe that in the future we'll be looking at even more electromechanical solutions--a good example is Otto Bock's C-Leg, which combines technology that monitors both position and prosthetic loading."
Stark foresees active movement of the lower-extremity prosthesis, including knees and ankles, rather than just passive compliance. "Virtually every company is looking at different ways of controlling the stance triggers they have--or looking at different types of control fluids. Recent examples are Biedermann-Motech's rheonetic fluid, and Ossur's stance control knee, TKO1500, which utilizes a cylinder type of braking system. And Ossur-Victhom is trying to provide active motion of the knee as opposed to passive resistance."
The lower-extremity prosthesis is primarily a passive device, Stark explained, so it is defined more by the movement it allows rather than the movement it initiates. By nature, in normal human locomotion, that's what walking is, anyway: a controlled fall.
"So," said Stark, "most of the motions we're using in normal human locomotion are eccentrically lengthening the muscle and just controlling the fall. A prosthesis is really just an extension of that--to aid in a controlled fall. The trouble begins when we encounter obstacles such as a hill or stairs. Now we've got problems, because now we require active movement of the knee to concentrically contract our muscles.
"One characteristic of lower-extremity prostheses is that we've never really had active movement as we do in the upper-extremity prostheses," Stark continued. "The tremendous load demands on the prosthesis and the weight constraints have always been prohibitive to adding extra gadgets and active movement."
More active motion is likely to be built into future legs, Stark believes. "Even the best prosthetic leg is a far cry from the anatomic leg and what it can do and how it adapts. And because we really don't have any active movement of the device in any way, I think we're going to be looking hard at things like that: how to pick up the toe, even using that motion to help generate some potential energy with a spring or as a trigger for electronic devices."
Looking far in the future, there will be a knee that moves as the result of a signal from the patient, Stark predicts. "In the near future, however, I think people are really looking at low-cost alternatives--trying to work with different mechanisms, fluids, and triggers to develop a better way to lock and unlock the knee when the patient desires it.
"When we look to the future, a lot of it is limited by the weight of the limb; and anything we make with motors and batteries has to be all self-contained," Stark continued. "We could do all kinds of wonderful things if we could pull behind them a wagonload of hydraulic controls--we could even make toes wiggle! But it's not going to happen in the near future because it has to be all self-contained.
"From a development perspective that's why we're all waiting to see what happens with osseointegration," Stark said. "With that attachment, the prosthesis could weigh a little bit more and we could add more gadgets and movement. The human leg weighs about 25-35 lbs. But we could never make a prosthesis that weighed that much-- the skin interface can't tolerate that much load, so we're shooting for seven to ten lbs."
Engineers could develop a lot more intricate devices if the patient could tolerate a little bit more weight, Stark explained. The problem has always been with the boundary layer, or the infection rate, he added, but with successful osseointegration, engineers can be more novel in their approach.
People did pretty well with prostheses in the past, Stark pointed out. "I think sometimes it's modern hubris that we assume that our stuff is so much better. In the 40s there are films of highly active patients playing basketball in quad sockets and exoskeletal legs. The componentry was relatively simply, but with effective use of alignment and interface design, high activity was possible."
Stark foresees greater control of microprocessors and perhaps even active movement, but added, "I think there will still be room for improved mechanical apparatus that are controlled in a more conventional manner."
Some prostheses from the past have been revisited with new technology, Stark noted. "An old hydrocadence knee was utilized in the 40s and 50s--and that was one of the few knees that actually allowed for toe pickup. I've heard developers talk about how significant it was, although the hydraulic system leaked and it was bulky. As such systems evolve; there might be ways to improve from mechanical solutions to hydraulic solutions, and now to electronic solutions!
"So basically, a lot of the new things aren't really that new," Stark continued. "Many of the concepts existed a long time ago, you can see in the literature of the time. The materials or techniques simply weren't present to make the design objectives possible or the devices consistently successful. When you look to new designs, I think we all try to revisit those products because of the advantages associated with them and attempt to improve on them with modern materials and design technique."
Charles W. Radcliffe, MS, ME, professor of mechanical engineering (emeritus), University of California at Berkeley: "There will be a continued effort to incorporate microprocessors into the control systems for artificial knees. Current designs are extremely simple devices compared to the possibilities of the engineering field of mechatronics.' Engineers in school today are eager to design components that will benefit amputees, and I am sure they will create designs far better and less expensive than what is available today."
Radcliffe added a caveat: "Unfortunately, a less expensive device may not be greeted warmly by some prosthetists. There is often a tendency to fit a more expensive device, whether or not the amputee can use it effectively. This is due to many factors, one being the amputee who wants the best' and believes the best must obviously be the most expensive."
The first application of microprocessors was to control the needle valves in a pneumatic swing control used with a single-axis knee, Radcliffe noted. "Current pneumatic swing controls with simple mechanical needle valves, properly designed and adjusted, will do a good job in controlling the speed of walking without a change in valve position after the valves have been adjusted for a particular amputee." Microprocessor control of the extension and flexion resistance valves may improve the response of the system, but the improvement may not be significant, he said, adding, "Current pneumatic and hydraulic swing control devices are more than adequate for all but very active amputees.
"The major benefit for microprocessor control is stance phase control using hydraulics or electromagnetic fluids," Radcliffe continued. "The control system components can be quite inexpensive." The system is typically assembled from components easily available and used by engineering students, he explained. The more challenging problems are in the design of special transducers to provide input of forces, moments, position, velocity, and acceleration. Then, with the possibility of many inputs, which set of inputs is optimal and what functions actually need to be controlled must be considered. "At present, the angular position and angular velocity of a single-axis knee are being controlled by the microprocessor system to provide both swing and stance control with a variety of sensors measuring control inputs. In the future there will be many improvements in the inputs and control strategies used."
In what direction would you PREFER to see future prosthetic knees evolve--and why?
"There is a tendency in prescribing and fitting transfemoral amputees with specified devices that provide knee security as the number one priority," Radcliffe pointed out. "As a result, there is widespread use of so-called safety' knees that provide automatic knee braking action at heel contact. As a result the amputee becomes a lazy' walker who depends on the device more than his own residual musculature.
"This may provide good security when walking on level surfaces, but may actually introduce a dangerous situation under some conditions, such as negotiating a down slope or coming down stairs where it may be difficult to unlock the knee under load. The amputee thus adopts a safe but slow one-step-at-a-time style of gait whenever a dangerous situation is encountered."
Hydraulic "swing and stance" control can provide stumble recovery by allowing the knee to flex slowly with hydraulic resistance, Radcliffe noted. This function is also theoretically useful for stair descent, but usually requires an athletic amputee to master the technique.
"I am a great believer in using the residual hip musculature on the amputated side as an excellent source of power and control for prosthetic knees," Radcliffe said. "This is particularly true for an active amputee using a four-bar knee. A properly designed and fitted four-bar prosthetic knee can provide excellent control of knee stability throughout the stance phase of the walking cycle.
"A four-bar knee, because of the kinematics of an elevated instant center, allows an effortless and energy-saving transition from stance to swing with control provided by hip muscles. The only missing function is stumble recovery. However, another feature of a well-designed four-bar knee is a definite increase in toe clearance during swing-through, hence the amputee seldom, if ever, stubs the toe and stumbles. Better voluntary control of knee stability also improves safety on rough ground and slopes. Unfortunately, few prosthetists really understand the function of four-bar knees and believe their only advantage is improved stability at heel contact. As a result, they are often aligned such that they function as a safety knee. This is a holdover from very early units that were specifically designed to provide a very safe knee."
Radcliffe would like to see a prosthetic knee that combines the flexibility of microprocessor control of a hydraulic device for the swing phase with the advantages of a polycentric knee in the stance phase. Such a device could be much simpler than current designs since the microprocessor would be used primarily for control of the resistance pattern in the swing phase, he said. A major cost item in current passive hydraulic swing control devices is the need for a mechanical system of adjustment of the flow through a series of very small holes. Microprocessor control of the resistance pattern is a simple problem of programming the resistance pattern of a single valve as a function of angular position and angular velocity of the knee, he explained, adding, "The program is easily varied to suit the needs of each individual amputee wearer. Sensor design and programming for stumble recovery is a much more challenging problem, and adds considerably to the cost of a microprocessor-controlled swing and stance device."
What advantages might mechanically enhanced knees offer that microprocessor-controlled knees could not?
"The primary advantage of mechanically adjusted devices is simplicity and low cost. These attributes do not necessarily mean poor function," Radcliffe said. Passive pneumatic devices have fewer leakage problems, he commented. "Hydraulic devices, with either mechanical or microprocessor control, do tend to leak eventually and require more frequent maintenance."
What Might Someday Be Possible?
Jules Verne taught us not to scoff at visionaries; what one man can dream, another can do. Many "science fiction" possibilities are being explored for prostheses that may be closer to market than we suspect.
Avenues being pursued include using animal tissue to create a "living knee" fed by nutrients rather than fueled by batteries. Some researchers are developing special actuators for knees that contain reactive gel. Joel Burdick, deputy director of the California Institute of Technology's Center for Neuromorphic Systems Engineering, Pasadena, is applying embedded chip technology that may allow the brain to directly control permanently attached prosthetic limbs. Reggie Edgerton, Life Sciences Department, University of California-Los Angeles (UCLA), is applying a similar philosophy to develop solutions for patients with severe spinal disabilities, who thus have difficulty walking.
And who knows what yet unimagined possibilities may suggest themselves--far sooner than we think?
Judith Otto is a freelance writer based in Holly Springs, Mississippi.