Adapting Current Prosthetic Stance Control Knee Joint Technology to an Orthotic Device:
A Preliminary Report
Brandon L. Khoury
Submitted in Partial Fulfillment of the Requirements
For an NCOPE Orthotic Residency
St. Louis, MO
May 2003
Abstract
The 2002 AOPA national convention unveiled major advancements in orthotic knee joint technology. Horton, Becker, Fillauer, and Otto Bock all presented their stance control knee joints. Four unique designs emerged and catapulted orthotic technology into the 21st century. Horton's Nassau knee was the first to arrive on the scene. This system utilizes a heel plate to activate the locking mechanism. Becker and Otto Bock both released similar designs that operate the lock with a cable system. Becker also has designs for the first computerized orthotic knee joint. Current available information indicates that the microprocessor will control both swing phase and stance phase, similar to Otto Bock's prosthetic C-Leg. The field of prosthetics has enjoyed stance control knees for many years. With this type of technology readily available, one may wonder why orthotics never adapted any of the designs. Could it be done? This study will compare and contrast the joint designs and locking mechanics of the currently available stance control knee joints with a prosthetic adaptation.
In the 1960s, prosthetic knee joints began a transition from single axis joints to polycentric. Four bar knee joints offered transfemoral amputees several characteristics that are inherent to their design: increased stability during stance phase, more anatomical motion, and improved cosmesis while sitting are three of the more favorable characteristics. Single axis joints achieve stability by maintaining the patient's weight line anterior to the joints axis of rotation. Four bar knees do not have a single axis of rotation, but many, referred to as the instantaneous axis of rotation (IAR). The inherent stability of these joints, during stance phase, comes from the initial location of the instantaneous axis of rotation. The axis is located superior and posterior to the anatomical knee center. To locate the IAR, simply extend lines through the anterior and posterior axes. The intersection of these two lines is the location of the IAR. The more superior and posterior to the anatomical joint, the more stabile the joint is during initial contact and stance phase. The initial IAR is key not only because of its stability but because it reduces the amount of energy needed to control the knee. The second advantage to polycentric knees was the motion it provides the patient. The human knee joint is commonly thought of as a hinge joint with one axis of rotation. This is, however, incorrect. The human knee joint is more complex than a simple hinge joint. When knee flexion is first initiated the joint begins to roll. The medial condyle rotates approximately 10° and the lateral 15°. After this rolling phase has been completed the joint begins to translate and rotate simultaneously. Since there is more movement of the lateral condyle, lateral rotation occurs. When approaching maximum flexion the joint only undergoes a sliding motion. The four bar linkage very closely mimics the anatomical joint by pivoting first, then sliding to complete flexion. It is this motion that also accomplishes the third before mentioned characteristic of improved cosmesis during sitting. At approximately 90° of flexion, the posterior/superior prosthetic axis falls inline with the two anterior axes. This position prevents the joint from protruding forward, decreasing the length of the prosthesis.
The next advancement was the addition of a 5th axis. The advantage to the additional axis is the incorporation of stance flexion. During normal gait the human knee is pre-flexed 5°-8° at initial contact. During limb loading the knee flexes to 17°-20°. Stance flexion occurs for two reasons: One, to act as a shock absorber of floor impact. Two, to help maintain forward progression. Knee flexion allows the center of gravity to maintain its' momentum forward in the sagital plane. The alternative would be a decrease of momentum to accommodate for the rise and fall of the center of gravity if the knee was to remain fully extended from heel strike to toe off.
The most recent advancement in prosthetic polycentric knee joints was the addition of two more axes. Developed by Century Twenty Two, the Total Knee was the first seven bars joint available. The joint offers all the advantages of the four and five bar knees but it also provides a geometric locking feature during stance phase. The joint locks itself to prevent knee flexion during single limb loading. The engagement and dis-engagement of the joint is dependent on three things; the position of the thigh over the lower leg, the location of the weight line, and the reaction to the ground reaction forces applied to the prosthesis. The following is a breakdown of how the seven bar knee reacts during each stage of stance phase:
Initial Contact:
The body's alignment at initial contact is 5°-8° of knee flexion, the foot is dorsiflexed, and the weight line is posterior to the knee axis. Biomechanically, this position would indicate knee flexion upon limb loading. The human body compensates by firing the knee extensors to prevent the knee from buckling. At this point an amputee, who lacks knee extensors, needs the knee to be fully extended as the heel contacts the floor. The patient does this by using hip extensors to apply a posteriorly directed force against the posterior socket as the heel contacts the floor. This, in effect, applies a rotational force to the prosthesis. The floor pushes anteriorly on the foot and the thigh pushes posteriorly on the socket causing rotation around the knee joint. The design of the knee prevents extension past 180°.
Loading Response:
During loading response the weight of the patient and the applied ground reaction forces, initiate the engagement of the locking system. As weight is applied, the ground reaction forces attempt to bring the foot to a neutral flat position. An anteriorly directed force is applied to the posterior/proximal aspect of the lower leg causing it to move forward. This force is accompanied by the firing of the hip extensors to initiate forward progression of the body over the limb. The combination of these two actions causes a posterior slide of the proximal linkage over the distal linkage around the pivot axis. As the posterior slide occurs, it causes the locking link to rotate forward and rest against a bumper. The weight line is now posterior to the pivot axis, causing knee flexion, and locks the locking link against the bumper. Remember, polycentric joints do not have a single axis but several. In this case one of the axes has been named the "pivot axis" because the surrounding links rotate around this axis, not because this is the pivot point of knee flexure.
Mid-Stance:
The lower leg and foot are at neutral positions and 90° to the floor. The thigh is directly over the lower leg but the hip extensors are still applying a posterior force against the socket wall to maintain forward progression. The knee joint is still locked because of the applied forces but the weight line is anterior to the pivot axis. At this point the neutral position prevents the shin section from moving forward and the knee acts similar to a single axis joint. With the weight line falling anterior to the pivot axis the joint wants to extend. The stability of the joint at this point is a combination of the patients weight maintaining the locking link against the bumper and the posterior/superior position of the IAR.
Terminal Stance:
Terminal stance is defined as the second half of single limb support. It is characterized by heel rise through the time when the contra-lateral foot contacts the ground. During this phase the weight line moves anteriorly, to the forefoot. This is the time when the joint lock is disengaged. As the heel rises from the floor, the ground reaction forces attempt to keep the foot flat. A posteriorly directed force is applied to the anterior/proximal section of the lower leg. At the same time the hip flexors fire, applying an anteriorly directed force on the anterior socket wall. This action is to maintain forward progression of the limb and initiate swing phase. The combination of these forces causes a forward slide of the superior joint linkage and a posterior slide of the inferior linkage. This rotation about the pivot axis disengages the geometric lock. The joint is now ready to flex during swing phase.
From Prosthetics to Orthotics:
Since the development of the first four-bar knee in the 1960s, Orthotists and engineers have attempted to adapt and apply many comparable designs. However, the only advancements in the last 40 years for people using KAFOs, was the availability of thermoplastic materials to replace metal and leather. The thermoplastic designs offer a decrease in brace weight and a more intimate fit. But, with many patients choosing to use wheelchairs, instead of KAFOs, a change of materials does not seem to be the answer. Why would patients, who have the ability to walk with a KAFO, choose to depend on a wheelchair? It is estimated that 60% to nearly 100% of all KAFOs are rejected after short periods of use. The most common reasons for discontinuing use were changes in the patient's needs, difficulty in obtaining the orthosis from the suppliers, unacceptable performance of the orthosis, and difficulty using the orthosis. But Kaufman et al hypothesized that the main reason for abandoning use was that KAFOs required too much energy. Most of the available research regarding gait efficiency was geared toward prosthetics. So, in the mid 1990s, Kaufman, Mathewson, and Sutherland tested the energy efficiency of locked KAFOs and a newly designed one that locked the knee during stance phase and allowed free flexion during swing phase. It was believed that the compensatory deviations in locked knee KAFOs were the cause of excessive energy expenditure. Such deviations include ipsilateral hip hiking, ipsilateral circumduction, and contralateral vaulting. It was hypothesized that the stance control KAFO would decrease the occurrence of these deviations, thus decreasing excessive energy costs. The subject tested in this study was a 40-year-old male with poliomyelitis in his left lower extremity. Manual muscle testing revealed 2/5 knee flexors, 1/5 knee extensors, and 1/5 plantar flexors. The subject was chosen because he used a locked knee KAFO on a daily basis and would be able to report valuable feedback. The results of the study did in fact support the hypothesis that the stance control knee joint decreased energy costs. But the joint not only decreased energy expenditure, it also decrease oxygen consumption and increased gait efficiency. However, the data also indicated that both knee joints resulted in higher costs and lower gait efficiency when compared to an able bodied individual. The results of this study, and similar studies, were the driving force for clinicians and engineers to develop functional, cost effective, stance control Orthotic knee joints. In the last several years Horton, Becker Orthopedic, Otto Bock and Fillauer all unveiled their own stance control joint designs.
 |
Picture 1. Horton Nassau Knee |
Horton: Nassau Knee
Though no documentation was found, Horton appears to have been the first company to commercially offer a knee joint that locks during stance phase and allows free flexion during swing phase. Gary Horton C.O. developed the Stance Control Orthotic Knee Joint (SCOKJ) over a period of eight years. The joint is a weight-activated system that utilizes a heel plate at initial contact. As the patient makes contact with the floor the heel plate pushes pins up into the locking mechanism by way of a cable system. The pins engage the lock preventing knee flexion during stance phase. The SCOKJ has the ability to lock at any degree of flexion. As the patient continues through stance phase the lock is disengaged at heel rise. With the lock disengaged the knee is ready to move freely through swing phase. The SCOKJ features three modes for ambulation. The joint can be switched from free knee to locked in 180° of extension or to the stance control mode. (See Picture 1).
Indications:
Bilateral quadriceps weakness/absence Unilateral leg paralysis/paresis Increased stability for GRAFO candidates Increased stability for offset knee KAFO wearers Increased stability for free knee KAFO wearers Increased stability for solid ankle/ Plantar Flexion stop AFO wearers
Contraindications:
Total loading greater than 225 lbs./102k Significantly impaired cognition Impaired balance Impaired motivation Uncorrectable genu varum greater than 10° Uncorrectable genu valgum greater than 10° Knee flexion contracture greater than 10° Biological knee joint greater than 5° off line of progression in swing phase
 |
Picture 2. Otto Bock Free Walk / Becker UTX |
Becker Orthopedic: UTX and Otto Bock: Free Walk
Becker and Otto Bock both released fully functional KAFOs with stance phase locks. The designs are the same and were developed by Dr. Nils Van Leerdam of Ambroise Holland. The brace is constructed of lightweight tubular steel. The entire KAFO weights less than 2 lbs. The locking mechanism is controlled by a cable system that is housed inside the tubular sidebar. A lever automatically locks the knee when the joint is fully extended. In order to unlock the joint, a combination of two actions must occur. First, an extension moment must be applied to un-weight the lever. Second, the ankle must dorsiflexed 10 causing the cable to pull the lever back away from the locking stop. With the lever clear, the knee is free to flex during swing phase. Though the system sounds complex the natural pattern of human gait will reliably engage and disengage the system. Both Otto Bock and Becker offer two braces. One is rated up to 175 lbs. the other is rated up to 265 lbs. (See Picture 2).
Indications:
Ipsilateral quadriceps weakness Polio Post-polio Multiple Sclerosis Unilateral paralysis Incomplete spinal cord injury Other traumas
Contraindications:
Knee flexion contracture greater than 10° Unstable valgus condition of the knee greater than 10° Unstable varus condition of the knee greater than 10° Severe spasticity Severe ankle instability Cognitive impairments
 |
Picture 3. Fillauer SPL |
Fillauer: Swing Phase Lock (SPL)
The most recent joint to be released was the Swing Phase Lock, or SPL by Fillauer. It was designed in the Netherlands, by Basko Healthcare. The joint is unique in that it is the only joint that does not require a footplate or a cable system to engage the lock. This system utilizes two joints that work together. The SPL joint was designed to be mounted laterally. The medial joint is the SPC, or Swing Phase Control. The SPL locking system is dependent on the angle of the joint in the sagital plane as the knee is extended. An internal pendulum engages a pawl to lock the knee. As the patient goes through stance phase the joint wants to flex, but is unable to do so. In order to disengage the lock the flexion loading forces must be removed. An extension moment must be initiated to disengage the pawl. If the patient is unable to reach the proper amount of extension, a dorsiflexion stop may be added to the brace to generate a ground reaction force sufficient enough to unlock the joint. The system also includes a satellite remote control. This allows the patient to select from three settings: automatic lock, manual unlock, and manual lock. The automatic lock is to be used for active ambulation. The manual unlock is to be used to aid in sitting or when the auto lock fails to engage. The manual lock is used to provide maximum stability. It is important to note that if the remote control should break, a security lock will engage. (See Picture 3).
Indications:
Post- Polio Spinal involvement Cardiovascular accident Peripheral paresis/paralysis Nerve inflammations Neurological failures Myopathies Multiple Sclerosis
Contraindications:
Knee flexion contracture greater than 10° Central paralysis Hip flexion contracture Hip musculature involvement Poor balance/coordination
Stage I-The Adaptation of the Total Knee:
The goal of this project was to successfully construct an orthotic knee joint that mimics the function of the Total Knee. Ideally, the actions that were previously mentioned about how the Total Knee functions, would be applied to the orthotic adaptation. This joint was chosen because it is a design that does not have to rely on a footplate, cable system or a complicated internal mechanism. The joint itself locks geometrically based on the location of the pivot axis, the weight line and the floor reaction forces. The construction of the Orthotic adaptation occurred in three phases: transfer of design, construction of functional model, and construction of functional prototype.
Transfer of design:
The initial plan was to exactly duplicate the specifications of the Total Knee. The materials that were chosen for the functional prototype led to slight modifications of the design. The first step in the transfer of design was to measure the dimensions of the Total Knee, length and angles between axes. This task proved to be difficult for two reasons. One, the design schematic was unattainable. Second, the only joints available to act as a model could not be disassembled. The measurements were taken from an assembled unit with a measuring tape and a goniometer. An inside caliper was used to measure the distance between the outside edges of two bearings. The width of a single bearing was then subtracted from the measured distance. The resultant number was the distance between the centers of the two axes.
 |
Figure 1. Axis numbering of the Total Knee:
#5- Pivot Axis, #6- Locking Axis |
The length measurements taken were as follows: (See figure for axis numbering).
1 to 2 1 to 3 3 to 4 4 to 5. 2 to 4 4 to 6. 6 to 7
Angular measurements were taken for axes that were adjacent to each other and shared a common link. The angular measurements taken were as follows:
3 to 4 to 5 (Numbers 1,2, &3 make up the triangular shaped link) 4 to 5 to 3 5 to 3 to 4 2 to 4 to 6 Mid line of lower upright to 5 to 7
Construction of the Functional Model:
From the measurements, patterns were drawn on construction paper and followed the contours of the Total Knee. The patterns were then transferred to 3/16" polypropylene. The only other materials used in the construction of the model were Chicago screws to act ass the axes. The links were cut out using a ban saw and ground by hand on a troutman. The links were then drilled and pieced together. The functional model worked properly but displayed two distinct problems. One, the M/L dimension of the superior/anterior links were too wide, approximately 15/16 of an inch. This would pose several problems in the functional model. The wearer of the KAFO would be prone to hitting the medial side of the brace against his contralateral limb. This could possibly cause injury to the sound limb and damage the joint. The other issue with the width of the joint would be one of cosmesis. It would be difficult for the patient to wear pants over a brace that has that type of bulk. The second characteristic that presented a problem was the "flimsiness" of the plastic components. This made it difficult to determine if the pieces to the puzzle would fit together correctly in a more rigid structure. The problem of the plastic flexing became most noticeable when the lock was engaged and flexion forces were simulated. At this point in time the plastic components flexed an extra 5°-10°.
 |
Picture 4. Limited flexion of Orthotic Adaptation Joint
(The joint is held together with 1/2 aluminum in pre-assembly) |
Before construction of the functional prototype began, the decision was made to alter the design of the joint. To help reduce the overall width of the joint, the two outer anterior/superior links were reduced to one and centered in the mid-line of the joint in the coronal plane. This change would give rise to another problem. By centering the anterior/superior link, the posterior/superior #2 axis would make contact with the anterior link. This would greatly reduce the range of flexion available. To help accommodate this new problem, the anterior link was bowed to allow the #2 axis to rotate inside of it.
This accommodation was not the ideal solution because it still limited flexion to 70°. But, it was considered acceptable for this stage of the project. (See Picture 4).
Construction of the Functional Prototype:
Aluminum was chosen for the links because of its strength, lightweight, and it's easy to cut and grind by hand. Chicago screws were again chosen to act as the axes but there was concern that they would not to be conducive for a well functioning joint. The construction of the prototype began by using the polypropylene models as a template for the aluminum version. The plastic links were clamped in place, traced and acted as a positioning jig for the aluminum axis bore. To ensure the spacing of the axes remained equal to that of the plastic model, the drill bit was placed inside the plastic bores and then driven through the aluminum. Again, this was done to ensure the exact duplicate of the plastic model. The next step was to cut the links using the ban saw and grinding them with a troutman. When the pieces were put together the superior and inferior uprights were not parallel to each other. They were approximately 45° to each other when the joint was fully extended. The before mentioned problem involving the Chicago screws was now evident. The screws and the aluminum structure created a high level friction which made the joint unacceptably difficult to operate.
Modification of the joint design:
At this juncture of the project it was necessary to adjust the design and components of the joint. The main goal was to drastically reduce the amount of friction within the joint without compromising the integrity of its' structure. Going back to the prosthetic design, the solution presented itself. The Total Knee uses pin roller bearings to create an almost frictionless environment. The idea at this time was to incorporate roller bearings into the aluminum links. The bearings selected were 1/4" inside diameter, 1/2" outside diameter and 3/16" in width. The inside opening was determined by the fact that 1/4" diameter stock bar was available in our lab. The 3/16" width was to match the 3/16" aluminum links. The 1/2" outside was the only diameter available with the other two. The materials that are now being used for the two joints are:
3/16" aluminum 42 roller bearings 1/4" diameter stock bar 28 "E" clips 56 joint washers 28 flat washers
 |
Picture 5. Fully extended Orthotic Adaptation Knee Joint
(The joint is held together with 1/2 aluminum in pre-assembly) |
Once the materials were collected the schematic of the joint had to be adjusted to accommodate the bearings. One idea was to simply increase all the dimensions by 1/3 of
their length. This would have made the joint unacceptably large. Instead, all new dimensions were calculated using the Total knee schematic as a guide. The overall look and function remained the same but the lengths and angles have slightly changed. (See Figure 2 & Picture 5).
The length measurements are as follows: (See fig. 1 for axis numbering).
1 to 2 = 1.5 in. 1 to 3 = 1.75 in. 3 to 4 = 1.5 in. 4 to 5 = 1.25 in. 2 to 4 = 1.75 in. 4 to 6 = 1.4375 in. 6 to 7 = .8125 in.
The angular measurements are as follows:
3 to 4 to 5 = 50° (Numbers 1,2, &3 make up the triangular shaped link) 4 to 5 to 3 = 80° 5 to 3 to 4 = 50° Mid line of lower upright to 5 to 7 = 30°
 |
Figure 2. The new design schematic accommodating the roller bearings. |
Discussion:
The completion of stage 1was marked by limited, unofficial, testing. The joints were mounted to a thermoplastic KAFO with double action ankle joints. The ankle joints were locked a 90° at this time. The knee joints were mounted with the #1 axis at the knee center. An able bodied subject donned the brace and attempted to walk with a normal gait. Immediately, three problems were observed. First, the locking mechanisms engaged each step without failure. Unfortunately, the joints failed to successfully unlock. Upon further inspection of the brace, it was determined that there were two causes of this failure. Most notably, the ankle joints were not truly locked at 90°. Though pins were placed in both the anterior and posterior channels, the attachment of metal joints to plastic allowed unwanted motion. The other source of the problem was in the positioning of the joint on the KAFO. Going back to the suggested alignment of the Total Knee, the #1 axis should be placed at the same height as the knee center. The m/l placement of the joint should allow the weight line to fall approximately 5 mm anterior to the #5 axis. For this unofficial testing the joint was placed nearly 2 cm anterior to the #5 axis. This placement would cause the joint to become excessively stable and prevent the unlocking of the knee. The second problem occurred when maximum flexion was reached. A major flaw in the design of the joint prevented the joint to properly return to an extended position from maximum flexion. During flexion the superior two axes rotate posteriorly around the pivot axis. Once maximum flexion has been reached, an anterior rotation of the two joints begins. This motion binds the links and prevents extension from occurring.
The third problem was the m/l instability of the joints. The instability of the joints was found in the system holding them together. The E clips did not adequately secure the links to prevent m/l movement. The combination of these three problems was determined to be enough to ultimately fail the function of the joint. Testing has been pushed back until these problems can be resolved.
Stage II- Testing:
The second stage of this project will begin in the near future. The testing of the joint will help determine the best position for the joint to be in to consistently engage and disengage the lock. The initial position of the joint will be based off the alignment criteria of the Total knee. This will act as the basis for comparison of other positions. The goal is to determine that the locking mechanism does in fact work and which arrangement best accomplishes that goal. 45 trial runs will consist of 20 paces each (10 steps per side). The number of times the lock successfully engages will be logged in the following charts. An additional set of charts will be used to log how many times the joint successfully unlocks. The best position for the joint will be the one that has a high success rate in each of the two categories
KAFO with a solid ankle: |
Trial 1 |
Trial 2 |
Trial 3 |
Trial 4 |
Trial 5 |
Knee Joint @ 90 |
|
|
|
|
|
Knee Joint @ 105 |
|
|
|
|
|
Knee Joint @ 85 |
|
|
|
|
|
KAFO with a 90°
plantar flexion stop
and free dorsiflexion: |
Trial 1 |
Trial 2 |
Trial 3 |
Trial 4 |
Trial 5 |
Knee Joint @ 90 |
|
|
|
|
|
Knee Joint @ 105 |
|
|
|
|
|
Knee Joint @ 85 |
|
|
|
|
|
KAFO with free
motion ankle joints: |
Trial 1 |
Trial 2 |
Trial 3 |
Trial 4 |
Trial 5 |
Knee Joint @ 90 |
|
|
|
|
|
Knee Joint @ 105 |
|
|
|
|
|
Knee Joint @ 85 |
|
|
|
|
|
Bibliography
Ayyappa, Ed. "Normal Human Locomotion, Part 1: Basic Concepts and Terminology." Journal of Orthotics and Prosthetics (1997): 10-17 Breakley, James and Marquette, Stuart. "Beyond the Four-Bar Knee." Journal of Orthotics and Prosthetics (1998): 77-80. Campbell, James. "Stance Controlled Orthotic Knee Joints: A Report on a Collaborative Research." Journal of Proceedings (2000) Free Walk: Stance Control Knee/Ankle System Otto Bock Health Care, 2002 Kaufman, K., Irby, S.E., Mathewson, J.W., Wirtha, R.W. and Sutherland, D.H. "Energy-Efficient Knee-Ankle-Foot Orthosis: A Case Study." Journal of Orthotics and Prosthetics (1996) 79-85 Stance Control Orthotic Knee Joint Horton Technology, Inc. The Swing Phase Lock (SPL) Chatanooga: Fillauer, Inc., 2002
|
|
