LIBRARY MIchIgan State 1 Untverslty PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or botoro date due. DATE DUE DATE DUE DATE DUE ,p ' .1 Rfidzmfl. MSU Is An Affirmative ActiorVEqual Opportunity Institution cmwa-o.‘ AN EVALUATION OF AN ABOVE-KNEE PROSTHESIS TO FACILITATE RAPID GAIT BY Kimberly Ann Lovasik A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Biomechanics College of Osteopathic Medicine 1993 ABSTRACT AN EVALUATION OF AN ABOVE—KNEE PROSTHESIS T0 FACILITATE RAPID GAIT BY Kimberly Ann Lovasik The desire and the need for an above knee amputee athlete to have a recreational prosthesis is evident when an estimated 10,000 young amputees with an active sports interest are considered. Conventional prostheses function well for slow controlled gait, but fail to meet the demands of rapid walking and jogging. Tests on an active 22 year old above knee amputee showed that stride length and stride frequency could be increased by changing the inertial properties of the lower portion of the prosthesis. A four camera, three dimensional motion analysis system was used to further investigate the effects of changing inertial properties on an above knee bench prosthesis. By adding either a one or 1.5 pound weight to various locations on the shank of the bench prosthesis, the effects during free swing were studied. Results indicated that the frequency of swing was increased with the 1.5 pound weight located ten centimeters below the knee joint of the above knee bench prosthesis. This condition also created less rebound effect and greater knee flexion of the prosthesis. All of these results indicated that changing the inertial properties of the lower portion of the prosthesis created the conditions necessary to permit rapid gait for an above knee amputee. DEDICATION I would like to dedicate this to three very special people who made the completion of this degree possible: To my parents, Robert and Patricia Lovasik, whose encouragement, support (psychologically and financially), and love, over the years have helped me to reach this pinnacle. To my brother, Steven Lovasik, who gave me big brother pep talks and the tools I needed to finish this project. How's this for a return on your investment? iii ACKNOWLEDGEMENTS My grateful acknowledgements to my committee members - 30> * Dr. Roger Rent for being such a wonderful teacher and source of encouragement over the past years. * Dr. James Rechtien for agreeing to be part of this on such short notice and his support. * Dr. Robert Soutas-Little for being my academic advisor and sharing his knowledge. You never cease to amaze me. big thank you to: LeAnn Slicer for always reminding me to smile and keeping me sane. Sharon Busch and Brenda Robinson for taking care of those important grad school details and deadlines I always forgot about. Special thanks to: *Kathleen Hillmer for assisting in testing and her encouraging words. Best Wishes for a successful future in your lab. Brock Horsley for his continued support and encouragement. Wishing you much success with the Human Kinetics Lab. David Marchinda for assisting with my research and always being there for me. James Patton for explaining the impossible (dpend) and continuously listening to me. Tamara Reid for...everything! You'll never know how much I appreciate all that you have done and continue to do... The University of Michigan had their Fab Five, I have mine. Thank you all for your friendship. It is truly a precious gift to me. Last, but never least... * Kevin John Palczynski - thank you for believing in my "infinite potential". Your love is a daily reminder of my happiness. With my heart, with my soul...always. iv TABLE OF CONTENTS PAGE LIST OF TABLES ........................................ vii LIST OF FIGURES ....................................... viii CHAPTER ONE SURVEY OF LITERATURE ................................ 1 ABOVE KNEE PROSTHETICS ........................... 2 RESIDUAL LIMB .................................... 4 COMPONENTS ....................................... 7 ABOVE KNEE PROSTHETIC RESEARCH - WALKING ......... 8 ABOVE KNEE PROSTHETIC RESEARCH - RUNNING ......... 18 CHAPTER TWO INTRODUCTION ........................................ 24 ABOVE KNEE AMPUTEE ............................... 24 OLYMPIC RACEWALKER ............................... 26 RACEWALK TRAINING ................................ 27 DOUBLE PENDULUM MODEL ............................ 28 ABOVE KNEE BENCH PROSTHETIC ...................... 30 CHAPTER THREE EXPERIMENTAL METHODS ................................ 33 TARGETS/MARKERS .................................. 33 AN OVERVIEW OF THREE DIMENSIONAL CALIBRATION ..... 34 CALIBRATION ...................................... 39 TESTING PROTOCOL ................................. 41 DATA COLLECTION .................................. 46 CHAPTER FOUR ANALYTICAL METHODS .................................. 50 DATA ANALYSIS - DATA COLLECTION PROGRAM .......... 50 AN OVERVIEW OF THREE DIMENSIONAL TRACKING ........ 51 TRACKING ......................................... S4 DATA ANALYSIS - DATA ANALYZING PROGRAM ........... 56 ANGLE CALCULATIONS ............................... 57 CHAPTER FIVE RESULTS AND CONCLUSIONS ............................. 58 RECOMMENDATIONS ....................................... 67 APPENDIX A - FIGURES 17 through 28 .................... APPENDIX B - FIGURES 29 through 40 .................... APPENDIX C - FIGURES 41 through 43 .................... APPENDIX D - FIGURES 44 through 46 .................... LIST OF REFERENCES .................................... vi 68 80 92 95 98 LIST OF TABLES TABLE PAGE 1 TEST CONDITIONS OF ABOVE KNEE BENCH PROSTHESIS... 42 vii LIST OF FIGURES FIGURE 1 STUMP SHAPE: CYLINDRICAL VERSUS CONICAL ........ 2 MECHANICAL FEATURES OF FOUR BAR POLYCENTRIC PROSTHESIS ........ 3 MECHANICAL FEATURES OF SIX BAR POLYCENTRIC PROSTHESIS ........ 4 TERRY FOX JOGGING PROSTHESIS .................... 5 DOUBLE PENDULUM MODEL ........................... 6 ABOVE KNEE BENCH PROSTHESIS ..................... 7 MOTION ANALYSIS COMPUTER SYSTEM ................. 8 VP320 VIDEO PROCESSOR ........................... 9 CALIBRATION STANDS AND SET UP ................... 10 BENCH PROSTHESIS START POSITION - FLEXED ........ 11 BENCH PROSTHESIS START POSITION - STRAIGHT ...... 12 TARGET PLACEMENT ON BENCH PROSTHESIS RIGHT SIDE VIEW ............... 13 TARGET PLACEMENT ON BENCH PROSTHESIS LEFT SIDE VIEW ................ 14 LINKAGE SYSTEM FOR BENCH PROSTHESIS ............. 15 TOE MOVEMENT PLOT FOR FLEXED START POSITION OF BENCH PROSTHESIS VERIFYING "BOUNCING" OF STRUCTURE ....... 16 EXAMPLE OF TEST CONDITIONS WITH KNEE EXTENDER ATTACHED AND NO WEIGHT... viii PAGE 12 15 22 29 31 35 36 40 44 45 47 48 55 60 61 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 KNEE AND HIP FLEXION AND EXTENSION ANGLES PLOT CONDITION - NO WEIGHT ................ KNEE AND HIP FLEXION AND EXTENSION ANGLES PLOT CONDITION - NO WEIGHT ................ KNEE AND HIP FLEXION AND EXTENSION ANGLES PLOT CONDITION - 1 POUND AT CENTER OF GRAVITY ..... KNEE AND HIP FLEXION AND EXTENSION ANGLES PLOT CONDITION - 1 POUND AT CENTER OF GRAVITY ..... KNEE AND HIP FLEXION AND CONDITION - 1 POUND 10 KNEE AND HIP FLEXION AND CONDITION - 1 POUND 10 KNEE AND HIP FLEXION AND CONDITION - 1 POUND 25 KNEE AND HIP FLEXION AND CONDITION - 1 POUND 25 KNEE AND HIP FLEXION AND CONDITION - 1.5 POUND AT CENTER OF GRAVITY.... EXTENSION ANGLES PLOT CM BELOW KNEE JOINT.... EXTENSION ANGLES PLOT CM BELOW KNEE JOINT.... EXTENSION ANGLES PLOT CM BELOW KNEE JOINT.... EXTENSION ANGLES PLOT CM BELOW KNEE JOINT.... EXTENSION ANGLES PLOT KNEE AND HIP FLEXION AND EXTENSION ANGLES PLOT CONDITION - 1.5 POUND 10 CM BELOW KNEE JOINT... KNEE AND HIP FLEXION AND EXTENSION ANGLES PLOT CONDITION - 1.5 POUND 25 CM BELOW KNEE JOINT... KNEE AND HIP FLEXION AND EXTENSION ANGLES PLOT CONDITION - 1.5 POUND 25 CM BELOW KNEE JOINT... CROSS PLOT OF CONDITION - CROSS PLOT OF CONDITION - CROSS PLOT OF CONDITION - CROSS PLOT OF CONDITION - CROSS PLOT OF CONDITION - CROSS PLOT OF CONDITION - KNEE ANGLE NO WEIGHT .......................... KNEE ANGLE NO WEIGHT. KNEE ANGLE 1 POUND AT KNEE ANGLE 1 POUND AT KNEE ANGLE 1 POUND 10 KNEE ANGLE 1 POUND 10 VS HIP ANGLE VS HIP ANGLE oooooooooooooooooooooooo VS HIP ANGLE CENTER OF GRAVITY ...... VS HIP ANGLE CENTER OF GRAVITY ...... VS HIP ANGLE CM BELOW KNEE JOINT.... VS HIP ANGLE CM BELOW KNEE JOINT.... ix 68 69 70 71 72 73 74 75 76 80 81 82 83 84 85 35 36 37 38 39 4o 41 42 43 44 45 46 CROSS PLOT OF KNEE ANGLE VS HIP ANGLE CONDITION - 1 POUND 25 CM BELOW KNEE JOINT.... CROSS PLOT OF KNEE ANGLE VS HIP ANGLE CONDITION - 1 POUND 25 CM BELOW KNEE JOINT.... CROSS PLOT CONDITION CROSS PLOT CONDITION CROSS PLOT CONDITION CROSS PLOT CONDITION ABOVE KNEE CONDITION ABOVE KNEE CONDITION ABOVE KNEE CONDITION ABOVE KNEE CONDITION ABOVE KNEE CONDITION ABOVE KNEE CONDITION OF KNEE ANGLE VS HIP ANGLE - 1.5 POUND AT CENTER OF GRAVITY ..... OF KNEE ANGLE VS HIP ANGLE - 1.5 POUND 10 CM BELOW KNEE JOINT... OF KNEE ANGLE VS HIP ANGLE - 1.5 POUND 25 CM BELOW KNEE JOINT... OF KNEE ANGLE VS HIP ANGLE - 1.5 POUND 25 CM BELOW KNEE JOINT... AMPUTEE KNEE FLEXION ANGLE PLOT - NO WEIGHT .......................... AMPUTEE KNEE FLEXION ANGLE PLOT - NO WEIGHT .......................... AMPUTEE KNEE FLEXION ANGLE PLOT - NO WEIGHT .......................... AMPUTEE KNEE FLEXION ANGLE PLOT - 1.5 POUND BELOW KNEE JOINT... ...... AMPUTEE KNEE FLEXION ANGLE PLOT - 1.5 POUND BELOW KNEE JOINT ......... AMPUTEE KNEE FLEXION ANGLE PLOT - 1.5 POUND BELOW KNEE JOINT ......... 86 87 88 89 90 92 93 94 95 96 97 SURVEY OF LITERATURE Locomotion is a complex phenomenon which is best described through a multidisciplinary approach. The most qualified approach comes from a domain of classical mechanics, more specifically known as biomechanics, that has the greatest responsibility for establishing the relevant scientific knowledge. Locomotion is the action with which something moves through space whether it be in the air, in the water, or on the land. It is achieved by coordinating movements of body segments employed in an interplay of internal and external forces (Cappozzo et al., 1976). The external forces represent all physical interactions between the body and the environment. These include the inertial, the gravitational, and the ground reaction forces. The internal forces are, among others, those transmitted by body tissues as muscular forces, tension forces transmitted by the ligaments, and forces transmitted through joint contact areas (Cappozzo, 1984). Humans depend on locomotion for a variety of daily activities. For most, walking is the form of locomotion utilized the majority of the time. However, as a fundamental motor pattern required in many games and sports, running is a prerequisite for satisfying participation and achievement in physical activities. Because of the steadily increasing emphasis over the last three decades on maintaining a healthy lifestyle, running has become a form of locomotion utilized in recreational 2 and/or competitive settings. Consequently, the ability to run at least short distances holds the key to a more active life for many individuals and carries with it potential physical, social, and psychological benefits (Enoka et al., 1982). Some of the benefits associated with running include cardiovascular endurance and weight control (Clausen, 1977; Scheuer and Tipton, 1977), stress management (Selye, 1974; Andersen, 1979; Farge et al., 1979), and enhancement of self—esteem (Ismail and Young, 1977; Morgan and Pollock, 1977). At the most basic level, running is also a "self- preservation" skill utilized in certain emergency circumstances. For most individuals, the transition from walking to jogging to running is smooth and effortless. However, for lower extremity amputees each of these cadences of gait present its own challenges (Smith, 1990). ABOVE KNEE PROSTHETICS In ancient times, small tree trunks shaped into peg legs were substitutes for human lower limbs. Today, mechanical substitutes have advanced to sophisticated prostheses with controlled motion. A widely accepted criterion for an above knee prosthesis is that it must have a normal external appearance and it must permit the amputee, without any undue mental or physical effort, to walk comfortably and safely at normal rates on level ground (Wagner and Catranis, 1954). In addition to these, a primary requirement of an above knee 3 leg prosthesis is that it enables the amputee to walk with a normal appearing gait pattern (Wallach and Saibel, 1970). As a person walks, the legs share the support function. One leg supports the moving body while the other swings forward in preparation for its next support role. .These periods of support and swing are referred to as the stance phase and swing phase of gait. An above knee prosthesis must provide support during stance phase to substitute effectively for the missing limb as well as control knee motion during swing phase to insure smooth entry into the next stance phase. The prosthesis should also give the amputee effective voluntary control over the starting and stopping of knee motion to insure the stability of gait, especially on rough terrain (Zarrugh and Radcliffe, 1976). Literature reveals that the desire to improve the rehabilitation of amputees has created a need for research in amputation surgery and prosthetics. Efforts have improved almost every area within amputee rehabilitation and amputees are being served better, however there will always be an opportunity to further improve lower extremity prosthetics (Stokosa, 1984). The advancement in research and new developments have impacted both prosthetic design and function, therefore a brief background in the process will be beneficial for this study. RESIDUAL LIME From a prosthetist's perspective, the ideal residual limb may be described as pain free, strong, well muscled, and as functional as the original limb as possible (Moore and Malone, 1989) The prosthetist also wants the stump to be as long as possible to maximize the amount of surface used in weightbearing during gait and to provide a longer lever for more effective control of the prosthesis (Stokosa, 1984). The muscles should have distal insertions and be able to contract in their antagonistic manner without pain. The skin should be smooth and conforming with natural tension, have a good vascular supply, have moveable underlying subcutaneous tissue, and have a well-healed surgical scar (Moore and Malone, 1989). Active muscles may hypertrophy and inactive muscles may atrophy, depending on their physiological composition, whether they have distal insertions, and to a larger extent, the amputee's amount of physical activity and desire to maintain muscle density (Moore and Malone, 1989). The residual limb is extremely important in the fitting of the prosthesis and the stump shape that provides the best function is cylindrical in shape as opposed to conical (Figure 1). The cylindrical shape allows for a greater area of distribution for the body weight load and helps in rotational control of the prosthesis (Stokosa, 1984). However, not all prosthetists agree with this. Some believe that there is no single geometric configuration that fits ~ a i i @ Figure 1: Stump shape: cylindrical versus conical I U (Example shown is a below knee stump) 6 most amputees within a particular amputation level (Moore and Malone, 1989). Anatomy is unique to each individual and it alone should determine the optimal shape to permit maximum comfort and function. Comfort and functional utilization of a prosthesis can be achieved only if the fitting of the prosthetic socket to the amputation stump is done well. The amputation stump, along with the socket, forms a lever to control the prosthesis during the swing and stance phases of the walking gait. The stump must also transmit the entire weight of the body to the prosthesis (Stokosa, 1984). If a socket fits the stump accurately, there will be greater comfort in the fit and greater efficiency in the use of the prosthesis. There are many steps that go into the design, the fitting, and the fabrication process for an above knee prosthesis. The overall process may take from ten to twenty appointments with a prosthetist depending on the complexity of the case, the number of socket fittings needed, and the number of different components tested (Moore and Malone, 1989). For this study, the specific steps taken to fit and fabricate the prosthesis are not necessarily important and will not be discussed. However, the components attached to the socket in the fabrication process are of importance and will be discussed in the next section. COMPONENTS Appropriate components for the above knee prosthesis, such as the foot and/or ankle, and the knee, are attached to the socket and aligned under the amputee's center of gravity for the amputated limb. There are more than twenty different prosthetic feet, and an approximately equal number of prosthetic knees, commercially available (Moore and Malone, 1989). A relatively new procedure in analyzing the capabilities and deficiencies of prostheses allows for an amputee to try different components and wear-test them both in the rehabilitation facility and in the amputee's own environment and daily activities. Sometimes, an amputee will report greater stability, smoother motion, and more confidence with prosthetic components other than those predicted to be the best (Moore and Malone, 1989). This is why testing a variety of components is so important. What may be unobservable or go undetected by the prosthetist's eyes, such as socket design, component function, alignment, efficiency, or weight, may be detected by the amputee's proprioceptive and kinesthetic abilities. Knowing this, one can see that the best method of assessing prosthetic fit is by keeping a line of communication open between the amputee and the prosthetist that allows for comments and suggestions to make appropriate adjustments. Many new developments in prosthetic feet are thought to be based on energy storage. This is better described as the ability of the foot piece to store gravity generated 8 energy - potential energy - when the amputee shifts the body weight to the foot which in turn deflects and compresses the prosthetic material. In theory, this stored energy is then progressively returned to the amputee by a rebound action - kinetic energy — as the foot moves through toe off (Moore and Malone, 1989). The transfer that occurs as potential energy shifts to kinetic energy helps to produce a more natural walking gait pattern; New developments in prosthetic knees focus on a number of different concepts. The knee component is designed to achieve greater stability in stance phase and greater variability in controlling the foot and lower limb during swing phase. Two other aspects include keeping the component light weight and allowing for alignment adjustability (Moore and Malone, 1989). Some knee components, offering stance phase stability, make it mechanically difficult if not impossible, for bending of the knee to occur when the amputee puts weight on the prosthesis. Other knee components, such as pneumatic and hydraulic systems help to achieve greater control during swing phase of walking gait. Newer designs incorporate a polycentric design which imitates anatomical knee action (Moore and Malone, 1989). ABOVE KNEE PROSTHETIC RESEARCH - WALKING Over the years, many gait studies of amputees have been conducted and each varies as to the objectives and the 9 methodology. When reviewing the literature for this thesis, it became obvious that the majority of the research has focused on below knee amputees as opposed to above knee amputees. The studies on below knee amputees, although informative and useful, will not be discussed. Instead, attention will be focused on a review of literature from some important studies on above knee amputees. Amputee gait has a higher degree of variability than normal gait because it is influenced by many factors that do not affect normal walking, such as level of amputation, prosthesis alignment, and socket fit. This variation has prevented the development of generalized gait models for amputees (Zarrugh and Radcliffe, 1976). Therefore, emphasis in studies has been placed on the understanding of "normal" gait patterns to establish the analytical basis and kinematic standards required to solve the specialized problems posed by pathological gait, such as those by an above knee amputee. Clinically, it has been noted that above knee amputees walk more slowly than normal. This is true for both the adults (Godfrey et al., 1975; James and Oberg, 1973; Murray et al., 1980) and for children (Boy at al., 1982). It has also been shown that a prolonged swing phase in the prosthetic limb is responsible for the slower walking speed (Drillis, 1958; Leavitt and Zuniga, 1973; Murray et al., 1983). In normal walking, two major muscle groups control the 10 swing phase of walking - the quadriceps and the hamstrings. The prosthetic leg without these two groups of muscles loses its swing phase control at the knee joint and behaves like a pendulum, causing unnatural gait (Patil and Chakraborty, 1991). Early designs by Murphy (1964), solved the problem of controlling the leg during swing phase using a design that provided frictional resistance at the knee joint. Lewis (1965) later introduced a hydraulic control system. Research then centered on improving the knee hinge mechanism to accelerate the swing phase of gait (Wallach and Saibel, 1970; Zarrugh and Radcliffe, 1976). Wallach and Saibel (1970) investigated the performance of a mechanical hydraulic mechanism used to control the motion of the shank relative to the thigh. They used a DUPACO 'Hermes' hydraulic control unit as a model for their study. This device used a combination of spring force and fluid damping to control motion. Springs at one end or both ends of the piston stroke, aided in reversing the motion of the shank without expending excess energy. Fluid damping was generated by forcing incompressible fluid through a series of staggered holes in the cylinder. The holes were staggered to provide a damping force which is a function of the relative position and relative velocity of the piston (Wallach and Saibel, 1970). Comparing experimental data gathered from normal walking patterns for non-amputees, they were able to develop specific performance criteria for an above knee prosthetic control mechanism. This criteria 11 included the determination of linear force displacement and power requirements necessary to duplicate normal walking gait. In the study conducted by Zarrugh and Radcliffe (1976), the swing phase of an above knee prosthesis was investigated. The model discussed simulated a prosthesis equipped with a four-bar polycentric knee joint and a pneumatic swing phase control unit (Figure 2). The simulation allowed for the prediction of the motion of the shank-foot combination as it was guided by prescribed hip and thigh trajectories and controlled by the prosthetic knee during the swing phase of walking gait. Knowing that the ability to control knee stability by hip joint musculature is influenced by the height of the prosthetic knee center above the floor, they used a specialized mechanism. First developed by the Biomechanics Laboratory at the University of California (Radcliffe and Lamoreux, 1968), this polycentric knee mechanism had a center of rotation that started high to fulfill the stability requirement and then descended rapidly with knee flexion to preserve the normal appearance of the prosthesis in sitting. A high knee center or instant center for thigh-shank rotation, provides the amputee with improved leverage for voluntary control of knee stability (Radcliffe and Lamoreux, 1973). Because the design of the above knee prosthesis involves many basic functional specifications, evaluating the stability features of different knee mechanisms or 12 Rubber extension stop Pneumatic swing control unit Prosthetic shank Figure 2: Mechanical features of four bar polycentric prosthesis (Zarrugh and Radcliffe, 1976) 13 assessing the performance of possible swing phase control units can be accomplished without actually constructing a prototype. By simulation, the effects of changes in the geometric or inertial properties of the prosthesis can be assessed. Once a general type of joint mechanism is selected, the model can be used to study the kinematics and parameters altered to give the most desirable result. The swing phase controller used in the study conducted by Zarrugh and Radcliffe (1976) simulated the action of the muscle groups that normally act about the knee joint. Immediately after toe-off, the controller action mimicked the quadriceps group by restraining the flexion of the shank and limiting toe rise above the ground. Before heel contact, the controller provided a hamstring-like action to reduce the angular momentum of the shank to prevent abrupt impact and subsequent rebound at full extension. The knee angle and velocity curves documented in the study emphasized the angular momentum characteristics of the above knee prosthesis. The curves showed that the model adequately reproduced the controller action by bringing the prosthesis into full extension with diminishing relative knee angular velocity (Zarrugh and Radcliffe, 1976). As predicted on the basis of analysis, the behavior of the prosthesis coincided with experimental data and future designs of above knee prosthetics would benefit from their research. Mechanical friction as well as hydraulic and pneumatic 14 systems have been developed to control the movement of the shank during swing phase. A study comparing hydraulic knee systems to constant friction knee systems found that the hydraulic knees resulted in a walking speed and shank swing phase that was closer to normal than the constant friction knees (Murray, 1983). Unfortunately, both mechanisms had their disadvantages. In a prosthesis that provides mechanical frictional resistance, natural gait cannot be produced and the leg becomes damaged due to wear. Generally, the hydraulic control system can provide a better control of the swing phase, but due to its own weight, the prosthesis becomes quite heavy (Patil and Chakraborty, 1991). However, when a pneumatic control system is provided at the knee joint, both of the disadvantages of the other systems can be eliminated and a better walking gait can be achieved (Patil and Chakraborty, 1991). In the same study mentioned above, Patil and Chakraborty (1991) analyzed a new polycentric above knee prosthesis with a pneumatic swing phase control (Figure 3). This research was initialized because the four-bar linkage mechanism used by many previous designers, for the knee joint of an above knee prosthesis, provided knee rotation suitable for walking, but not for squatting. Conducting their research in India, they recognized the need of people in Afro-Asian countries to have a prosthesis that would allow them to sit in the squatting position used for many daily activities ranging from farm operations to daily 15 I I J ,{ mp SI'UIP s msmems rates we»: met: csmae // / l I art-sumo" m . tocx Farm was ,4? UNKAGES CONNECTING (9) men ~ We 'i SLOT moses W euros PLATE “3 BM - 2': ° LE em serum -2 Figure 3: Mechanical features of six bar polycentric prosthesis (Patil and Chakraborty, 1991) 16 prayer. By adding two linkages to the four-bar linkage mechanism previously discussed, they developed a six-bar linkage knee-ankle mechanism designed to accommodate the squatting position. Results of their analysis showed that the six-bar linkage mechanism was able to provide a near normal squatting position and that motion patterns of the ankle joint and the instantaneous knee center obtained by analysis during simulated walking, matched those of an amputee wearing the new prosthesis. Also, the linear displacements and angular movements of the hip, knee, and ankle during swing phase of the above knee amputee walking gait corresponded to patterns of normal walking gait. It was clear that the new polycentric above knee prosthesis provided functional improvement over the earlier developed four—bar polycentric unit. In other research investigating swing phase control of an above knee prosthesis, it has been hypothesized that by adjusting the friction in the knee unit, the pendulum action of the shank can be set to correspond to the opposite limb and enable the amputee to walk at a faster cadence (McCullough et al., 1981). However, if the shank and the foot of the above knee prosthesis is considered as a pendulum with distributed mass and a constant friction joint, the solution of the second order nonhomogeneous differential equation for sinusoidal motion gave the period of: 17 I T=2 (l— 7 Myd where T is the period, I is rotational inertia of the leg/foot about the knee joint axis, M is the mass of the leg/foot, g is acceleration due to gravity, and d is the distance from pivot to center of mass (Hicks et al., 1985) Based on this formula, one can see that the knee friction magnitude would have no effect on the natural period of the shank/foot - it would affect only the amplitude of the range of motion. Therefore, greater applied torque and increased energy consumption would be required for an amputee with a constant friction knee joint to walk with a cadence significantly different from a natural cadence (Hicks et al., 1985). Using this knowledge, Hicks and colleagues conducted tests on the effect of knee friction adjustments within large limits on swing phase dynamics in young, average age 10.5 years old, above knee amputees. Their data confirmed that above knee amputees walked slower than normal and it was the prolonged swing phase of the prosthesis that was responsible for the slower cadence. They also found that regardless of the amount of friction applied to the knee, the swing phase on the prosthetic limb took the same amount of time. One of the most interesting 18 findings was that there was no significant difference between the shank period as measured dynamically with various frictional settings and the period of the shank modeled as a pendulum. Hicks and colleagues concluded that efforts to decrease the prosthetic swing phase time would be better directed at decreasing the periodicity of the shank rather than at the knee mechanism. Referring back to the equation for the period of the sinusoidal motion, it is apparent that altering the inertia of the shank should alter the period. With an experimental prosthetic, they were able to alter the inertial properties by redistributing mass. They tested the prosthesis with the center of mass located 18.7 cm from the knee axis and then 31.7 cm from the axis. The period was approximately .20 seconds faster with the mass attached proximally. From these results, they concluded that the slower walking speed of above knee amputees, causing the prolonged swing phase, might be altered by redistributing the mass of the leg so that the center of mass is more proximal than that of a conventional prosthesis (Hicks et al., 1985). ABOVE KNEE PROSTHETIC RESEARCH - RUNNING From all the literature reviewed in the previous pages it is obvious that there has been considerable research directed toward studying and improving the walking gait of amputees. However, little effort has focused on the investigation of their running capabilities. As was 19 mentioned in the very beginning of this thesis, running has many potential physical, social, and psychological benefits (Enoka et al., 1982). In fact, exercise in general has many benefits as well as being an important aspect of rehabilitation. All physical activity subjects amputees to mental and physical challenges, although recent progress in developing programs for the physically handicapped has started to change the stereotypes that the handicapped must be inactive. Bone cancer has been the primary cause in the increase in numbers of young amputees. In a study conducted by Kegal and colleagues (1978), many of these amputees found distance walking and running difficult, but they still expressed a desire to run. Many of the amputees questioned were active in sports before losing a limb and were determined to continue despite their disability. In a statistical analysis conducted by Kegal and associates (1980), it was found that the ability to run at least short distances was the key to a more active life for lower extremity amputees. Many of the individuals indicated that the skill was difficult for them, as they did in the earlier study by the same author. Running is a basic locomotor skill incorporated into recreational activities as well as in emergency "flight" situations such as avoiding moving vehicles. Therefore, running performance of lower extremity amputees is important from a quality of life standpoint and for safety reasons (Kegal et al., 1980). One other finding 20 from the statistical analysis was that active amputees wanted recreational protheses. Unfortunately, recreational protheses are not covered by some insurance companies and can cost up to $200 more than conventional prostheses. A majority of active amputees indicated that cost was not the determining factor in acquiring a recreational prothesis, instead it was the prosthesis itself that discouraged them from acquiring one. Little research has been done on improving the design of these specialized prostheses despite the high demand from lower extremity amputees, especially above knee amputees. However, there has been a recent study which carries importance in changing this trend. A study by DiAngelo and colleagues (1989) reiterated that active above knee amputees need a recreational prosthesis that allows them to jog, that provides a natural, balanced, prosthetic to anatomical leg stride, and finally, that enables them to actively participate in sports. Although acceptable for walking gait, conventional prosthesis have failed to meet the demands of rapid walking and jogging. So with the help of Chedoke—McMaster Hospital in Hamilton, Ontario a jogging prosthesis was developed and assessed. Terry Fox, the late above knee amputee cancer patient famous for his 1980 fundraising run across Canada with Steve Fonyo, proposed an idea that replaced the metal shank of his conventional prosthesis with a pogo stick. Later, DiAngelo and colleagues collaborated with Chedoke-McMaster Hospital 21 to develop what was called the "Terry Fox jogging prosthesis" (Figure 4) which incorporated a large compression spring in a telescoping shank that compressed during jogging, similar to the way non—amputee joggers flex their knee during weight acceptance, and therefore, lessened the impact force. The energy stored in the spring was released at toe-off to help the runner propel forward and upward (DiAngelo et al., 1989). Results of the biomechanical assessment of the jogging prosthesis (DiAngelo et al, 1989) indicated that there was an increased swing phase time for the prosthetic limb and that the knee was hyperextended throughout stance phase and exceeded the normal level of knee flexion during swing phase. Also, ground reaction forces were almost double those of a non-amputee jogger. The assessment of the jogging stride of an above knee amputee wearing the "Terry Fox jogging prosthesis" was conducted at the University of Waterloo in Canada. The amputee tested with the jogging prosthesis was asked to jog at a comfortable pace and adopted a style of gait which consisted of alternating periods of single support and non- support for both the amputated and anatomical leg. When above knee amputees attempt to jog with a conventional prosthesis they cannot produce the rapid extension of the lower limb to a hyperextended, locked position to prepare for heel strike and weightbearing. Because of this, the jogger must wait for the prosthetic knee of a conventional 22 r, OPENS-END ' : SUCTION SOCKET O» TEH LIN FOUR BAR 7 PNEUMATIC KNEE u. ._ ‘“SHIN-TUBE 'SPRING UNIT GRIESSINGER FOOT Figure 4: Terry Fox jogging prosthesis (DiAngelo et al., 1989) 23 prosthesis to lock and then must vault over the straightened leg which produces the unusual jogging pattern of double support periods on the anatomical leg (DiAngelo et al., 1989). Although the jogging prosthesis reduced the vaulting effect commonly seen with an above knee amputee using a conventional prosthesis for jogging, the spring did not provide enough energy needed to propel the jogger forward at toe-off nor did it reduce loading. The prosthetic loads transmitted to the stump-socket interfaCe and the residual hip joint were approximately 4.3 times body weight. Loading conditions of this magnitude are unacceptable considering normal running loading conditions are approximately 3 times body weight. DiAngelo and colleagues made some breakthroughs in a design for a recreational prosthesis, but the Terry Fox jogging prosthesis still did not successfully satisfy all the requirements necessary to permit rapid gait. INTRODUCTION The desire and the need for an above knee amputee athlete to have a recreational prosthesis is evident from the studies mentioned in the survey on the previous pages. The "Terry Fox gait", or style of jogging which places an intermediate hop on the anatomical limb while the prosthetic limb is in swing phase, is the jogging style used by many active above knee amputees. However, it is also evident from one study (DiAngelo et al., 1989) that this is an unacceptable style of gait due to the loads put on the anatomical limb and the prosthesis as well as the prolonged amount of time it takes for the prosthesis to swing through during running gait. Research for this thesis utilized prior data collected from human subjects, an above knee amputee and an Olympic racewalker, at the Biomechanics Evaluation Laboratory in East Lansing, Michigan. Prior to testing any subject, the testing procedure was explained and required informed consent forms were obtained from each subject (UCRIHS 89—559 approval). ABOVE KNEE AMPUTEE The above knee amputee was a 22 year old male who had his left leg amputated four inches above the distal end of his femur after being diagnosed with cancer at the proximal 24 25 end of his tibia at the age of eleven. He was fitted with his first prosthesis three months after surgery and over the course of ten years he had been fitted for many types of prostheses, each time trying new components and taking advantage of new developments in the field of prosthetics. His first attempt with a sports oriented prosthetic system was playing goal keeper for various soccer teams. He was an avid athlete prior to his amputation and wanted to continue to remain active. In the spring of 1989 he began serious training for triathlons with the encouragement of an engineering student, and friend, at Michigan State University. He was successful in the swimming and cycling portions of a triathlon, however, it was the running portion that would set him far back from the rest of the competitors. To run, the above knee amputee triathlete adopted the "Terry Fox gait" using the alternating hop and jog running style as was previously described. Preliminary tests conducted at the Biomechanics Evaluation Laboratory analyzed this style of jogging. Using an AMTI force plate, ground reaction forces were investigated. One of the most notable findings from the analyses was that loading during the intermediate hop on the right, or anatomical, side was over four times body weight. Normal running forces are approximately two and one half to three times body weight. The repeated high loading condition experienced by the amputee during the running phase of a race, posed a risk of degenerative joint disease on the non-amputated side. These -26 tests showed that the "Terry Fox jogging gait" was an inefficient and stressful running form and not conducive to the above knee amputee attaining his goal of competing in a triathlon. OLYMPIC RACEWALKER An Olympic racewalker was tested to obtain baseline data on the sport. The gait employed in racewalking is a function of the rules which govern the sport. In order to execute a legal gait, constant contact with the ground must be maintained at all times, and a double support period must be established during each cycle. The supporting leg is also required to be "straight in the vertically upright position" (IAAF Handbook, 1984). Racewalkers tested by Cairns and colleagues (1986), demonstrated a significantly greater maximal extension to a position of hyperextension during stance phase, as opposed to walking or running. Also, ground reaction forces measured in that study indicated that racewalking loads were less than that of running. The tests performed at the Biomechanics Evaluation Laboratory confirmed the ground reaction force data from the Cairns et al. study, indicating that ground reaction forces were under two times body weight at a racewalking pace of 3.8 to 4.4 meters/second. It became obvious that the requirements of racewalking matched the functional limitations of the above knee amputee who could not generate a flight stage, who had to have the prosthetic 27 knee locked during stance phase and had the anatomical knee fully extended during swing phase to allow the prosthetic limb enough room to swing through during gait. It was decided to train the above knee amputee in racewalking and to adjust his prosthesis accordingly. RACEWALK TRAINING To increase the speed of the amputee's walking gait, the stride length and the speed of the gait needed to be increased while the double stance time during the transfer of weight from one limb to the other needed to be decreased. The double stance phase, no matter how short, eliminates a flight stage, and differentiates racewalking from running. It was felt that the above knee amputee with a leg prosthesis cannot generate enough power on the prosthetic side to have a flight stage. However, while a flight stage is theoretically possible from the anatomical leg to the prosthetic leg, the amputee cannot functionally tolerate the forces generated on the prosthesis. The amputee was instructed, in a 1990 pilot study at the Biomechanics Evaluation Laboratory, to try to increase his stride length and walk as fast as possible. At a pace of approximately 1.5 meters/second, the prosthesis had a prolonged swing phase and did not reach full extension at heel strike. The amputee's gait appeared to be "choppy" and unsmooth at that pace. In subsequent racewalk training sessions, a 1.5 pound weight and a 2.5 pound weight were added to the shank of the 28 prosthesis at various locations. Gait speed increased to approximately 2 meters/second with the 1.5 pound weight attached just below the knee on the prosthetic shank and the amputee expressed confidence in "feeling the leg snap through." His gait became much more smooth as compared to the test conditions with no weight on the shank. Analysis of the kinematic data from these racewalk training sessions showed a greater amount of knee flexion on the prosthetic leg with the 1.5 pound weight attached proximal to the knee as compared to knee flexion angles with no weight or 2.5 pound weight attached to the shank. From the initial indications, changing the inertial properties of the prothesis allowed gait speed to be increased, greater flexion of the knee, and subjectively, the amputee's gait appeared to be smoother and more "natural". DOUBLE PENDULUM MODEL Subjective data from the above led to the creation of a mathematical model for an above knee prosthesis. Mr. James Patton, a graduate research assistant at Michigan State University, developed a computer program to allow the change of parameters of a double pendulum. The above knee prosthesis was modeled as a double pendulum with lengths a, b, and c and masses m1 and m2 (Figure 5). Each of these parameters could be changed and effects of these changes evaluated. Increasing values of m2 caused the frequency to increase, and therefore the period to decrease because of 29 /////// {/ // // k' \ Figure 5: Double pendulum model 30 the inverse relationship between frequency and period. Increasing the value of c caused the frequency to decrease, or the period to increase. These mathematical findings coincided with subjective data gained from the above knee amputee. The double pendulum mathematical model does not duplicate the true motion of an above knee prosthesis for the following reasons: 1.) It allowed for free oscillation at the upper joint whereas the hip movement in a prosthesis is controlled by the amputee, and 2.) There was two-way oscillation permitted at the knee joint and in a prosthesis this is controlled by specialized mechanisms. However, the mathematical model did give insight into the idea that altering the mass of the shank could effect characteristics of an above knee prosthesis and it helped lead to the development of a structural model for an above knee prosthesis. ABOVE KNEE BENCH PROSTHETIC With the help of Wright and Filippis, Inc., a prosthetics, orthotics, and healthcare equipment company located in Rochester, Michigan, a bench prosthetic was modeled to simulate the leg of the above knee amputee (Figure 6). The bench prosthesis was able to swing freely from its supports and some parameters could be changed without redesigning the prosthesis. The experimental parameters which were altered for this thesis included the addition of a knee extender using surgical tubing and the 31 Figure 6: Above knee bench prosthesis 32 change of inertial properties by adding mass to the shank of the bench prosthesis. It was hypothesized earlier in the pilot studies, that changing the inertial properties of the shank on the prothesis, would permit the above knee amputee to attain rapid walking speeds and from initial tests this was true. This thesis investigates the possibility that, given the right prosthesis design, rapid gait is physically possible for the amputee. EXPERIMENTAL METHODS Data for this study was acquired at the Biomechanics Evaluation Laboratory, located at the St. Lawrence Health Science Pavilion, in East Lansing, Michigan. The laboratory is both a research and teaching facility and its technology includes: the measurement of three-dimensional motion, the measurement of the ground reaction forces which act on the body during motion, and the measurement of muscle activity utilizing a telemetered electromyography system. This thesis examines data using only the kinematic component of the three measurements listed above. TARGETS/MARKERS Spherical objects, such as ping pong balls and beads, were covered with 3M Scotchlite Brand High Grain 7610 Sheeting, commonly known_as retro-reflective tape. The composition of retro-reflective tape is small spheres that reflect light rays directly back to the original source. Unlike active targets, these passive cooperative targets contain no energy source, but instead work with an active external source. The retro—reflective targets were illuminated by four 100 watt indoor flood lights mounted next to each camera as closely as possible. To be effective, retro-reflective markers must be used with an illumination system that is in close proximity to the axis of the camera (Motion Analysis, 1990). This helps to reduce 33 34 the observation angle between the incident light ray, the retro—reflective target, and the reflective light ray. Also, when the incident light is placed coincident with the camera lens, the luminance factor is 1600 times brighter than a perfect white light diffuser (3M Product Bulletin). An increase of only one degree of the observation angle between the light source and the camera lens can decrease the intensity of light entering the lens by a factor of sixteen. Knowing this, one can understand the importance of the camera and light source placement and its effect on data collection and later, data analysis. AN OVERVIEW OF THREE DIMENSIONAL CALIBRATION The ExpertVision System by Motion Analysis, Inc., used for calibration and recording position data, consisted of four pixel perfect NEC high speed video cameras with a shutter speed of 60 Hertz and a VP320 video processor for digitizing data (Figures 7 and 8). Each frame of video data was synchronized in time for each of the four cameras. Inside the cameras, visual images from the lenses of each camera was focused on light sensitive surfaces. This produced electrical outputs sent back to the video processor. At this point, a threshold setting on the video processor set a line of demarcation between images that appeared either white or black on the display screen. Voltage differences between adjacent pixels crossing the threshold value were marked on the video screen as hollow 35 Figure 7: Motion Analysis computer system 36 Figure 8: VP320 video processor 37 targets. An ExpertVision software digitizing program then computed the centroids and determined two dimensional camera coordinates, for each hollow image (Motion Analysis, 1990). Three dimensional tracking, necessary for data analysis, is performed in two stages. First, each camera view is calibrated to establish the relationship between real positions, or object coordinates, and the corresponding image coordinates recovered from the individual views. Calibrations performed for two or more views, allow the researcher to employ the information generated during calibration to track one or more targets in three dimensional space. The tracking process will be described in greater detail in the Analytical Methods - Chapter Four. The calibration of a given view is specific to the configuration of the camera with respect to an arbitrary reference frame defined by the researcher. This reference frame is defined as the object reference frame. Some of the parameters which can influence the calibration are the position and orientation of the camera and the focal length of the lens, so any change which could alter the relationship between the object coordinates and the image coordinates would need to be followed with a new calibration of the space. The calibration process recovers values for eleven coefficients which implicitly define the configuration of a particular view. These calibration coefficients together with the image coordinates of a single target are sufficient 38 to define the path of an optical ray through the object space. If image coordinates are available for two or more views, and if each view has been calibrated, then the three dimensional path of an optical ray can be defined for each view. Under ideal conditions, using time-matched image coordinates, corresponding rays for a single target intersect at the location of the target. Therefore, the tracking process is one of intersecting optical rays generated from different views of the same event (Motion Analysis, 1990). In order to calibrate the available views, the researcher must first define an object reference frame. This is achieved by selecting an origin and three mutually orthogonal axes. All results rendered by the tracking process are referred back to this reference frame. The selection of an object reference frame is arbitrary, however, judicious selection will allow for easier analysis. The most important factor considered when choosing a reference frame for this specific study was the predominant direction of motion. For this thesis, it was more convenient to view the results as motion along the positive x-axis, so that forward motion of the bench prosthesis during free swing was from left to right. Once the reference frame was selected, a number of targets with known locations were used as control points. A minimum of six non-coplanar control points are required to calibrate a single view and the use of redundant targets tends to 39 minimize the impact of improperly located control points. It is recommended that for a good calibration, eight to sixteen control points be used (Motion Analysis, 1990). CALIBRATION Sixteen retro-reflective targets were positioned at known three dimensional positions to create a calibrated space. Four calibration stands were used to vertically suspend four targets per stand on a plumb line. Plumb bobs located at the end of each line allowed for more accurate placement of the control points over the four corners of a measured horizontal calibration area (Figure 9). Vertical measurements of the control pOints were made so that each represented a known three dimensional location in space. The calibration space used for testing and analyzing the bench prosthetic was 120 centimeters by 90 centimeters with vertical placement of control points in each corner at 25, 50, 75, and 110 centimeters. These dimensions assured that the predetermined bench tests would be performed within the calibrated space. For each camera, the two dimensional positions for each control point in pixel space were computed by the ExpertVision software digitizing program. Target numbers were assigned to the images in each of the four camera views and the ExpertVision software determined the three dimensional position of each control point. The "initializing" step in this process required the researcher 40 Figure 9: Calibration stands and set up 41 to manually identify each target with a number for each camera view. By direct linear transformation, the two dimensional information is combined from each camera to determine the perceived position of each target in three dimensional space. These perceived coordinate locations were compared to the measured coordinate locations. The degree of error between these two values was determined by a least squared program in the software (Motion Analysis, 1990). The camera location and pixel residual data were examined and the space was to be recalibrated if the norm of residuals exceeded .50. TESTING PROTOCOL The above knee bench prosthesis underwent a series of free swing tests changing a number of parameters. Weight conditions used were either no weight, one pound, or 1.5 pounds. Placement of the weight on the shank was at the center of gravity, 10 centimeters below the knee joint, or 25 centimeters below the knee joint. The center of gravity of the shank was grossly determined by detaching the lower portion of the prothesis and balancing it. The location of the weight placed on the shank was determined by measuring from the knee axis to the center of the weight. The knee extender rubber band built into the prothesis was either left unattached or it was attached according to instructions from the designer. Starting positions for testing were either flexed approximately 90 degrees at the knee or fully li2 TABLE 1 Test Conditions of the Above Knee Bench Prosthesis IGHT C0_DIII l8 LOCLIIOIS 0? IIIQEI Ill! BX!!!QII STARTING POSITIOI no veight III not attached flexed no veight Ilh attached flexed no relght III not attached straight no weight lid attached straight 1 pound center of gravity not attached flexed 1 pound center of gravity attached flexed 1 pound center of gravity not attached straight 1 pound center of gravity attached straight 1 pound 10 ca helox knee joint not attached flexed 1 pound 10 ca helav knee joint attached flexed 1 pound to ca helox knee joint not attached straight 1 pound 10 cu helov knee joint ‘ attached straight 1 pound 25 ca helor knee joint not attached flexed 1 pound 25 on below knee joint attached flexed 1 pound 25 on below knee joint not attached straight 1 pound 25 ca helov knee joint attached straight 1.5 pound center of gravity not attached flexed 1.5 pound center of gravity attached flexed 1.5 pound center of gravity not attached straight 1.5 pound center of gravity attached straight 1.5 pound 10 ca helov knee joint not attached flexed 1.5 pound 10 CI helov knee joint attached flexed 1.5 pound 10 ca helov knee joint not attached straight 1.5 pound 10 CI helov knee joint attached straight 1.5 pound 25 ca helov knee joint not attached flexed 1.5 pound 25 on below knee joint attached flexed 1.5 pound 25 CI helov knee joint not attached straight 1.5 pound 25 ca helov knee joint attached straight 43 extended. A breakdown of the tests can be seen in Table 1. The weights used for testing the bench prosthesis were ankle weights which were wrapped around the shank in appropriate positions and secured with self adhesive tape as well as athletic training tape. The tape did not alter the weight and because the design of the prosthesis using a flex foot attachment had long screws holding the prosthetic foot in place, the weight did not slide down the shank. This was verified in pretests by the researcher prior to conducting the tests used for this thesis. Graduate research assistants assisted in putting the above knee bench prosthesis in appropriate starting positions. If the starting pesition were with the knee flexed at 90 degrees, the assistant would hold the heel of the flex foot up so that the prosthetic femur was perpendicular and the prosthetic shank was parallel to the floor of the lab (Figure 10). If the starting position were with the knee extended, the assistant would hold the femur of the prothesis so that the prosthesis was straight and was at an angle approximately 45 degrees from the perpendicular (Figure 11). It was important for the assistant not to block any of the targets from the view of the cameras while holding the prosthesis in its starting position and to remain out of the view of the cameras once the testing began, so as not to interfere with data collection. 44 . 3) Figure 10: Bench prosthesis start position - ‘ flexed 45 : fl Figure 11: Bench prosthesis start position - straight 46 DATA COLLECTION To conduct tests using the above knee bench prosthesis, six targets were used. The targets were placed on the prothesis with double sided plastic carpet tape approximating what would be the bony landmarks of the hip, knee, heel, and toes on a fully fabricated above knee prosthesis. Two targets were placed on each side of the hip axis and knee axis of the bench prosthetic, one was placed in the middle of the heel region of the flex foot and one was placed in the middle of the toe region (Figures 12 and 13). The targets were used to define theoretical links and to approximate the motion of an above knee prosthesis for later analysis. After the targets were in place, the appropriately sized ankle weight was securely attached to the prosthetic shank and the knee extender rubber band was attached if the test condition called for it. The above knee bench prosthesis was then positioned within the calibrated space so that forward swing of the leg was in the positive x-direction of the previously defined reference frame. The research assistant was instructed to hold the prosthesis in the proper starting position while the researcher prepared the computer for the collection of data. The computer was triggered manually by the researcher to start recording data prior to when the assistant let go of the prosthesis, so that it would swing freely from its supports. Data collection automatically ceased after six seconds, a predetermined time amount chosen by the 47 Figure 12: Target placement on bench prosthesis right side view 48 Figure 13: Target placement on bench prosthesis left side view 49 researcher, and the trial was repeated. This process continued until two trials of each condition listed in Table 1 were complete. ANALYTICAL METHODS The computer programs used for data collection and data analysis at the Biomechanics Evaluation Laboratory were developed by either students or faculty associated with Michigan State University. The exception to this is the ExpertVision three dimensional software program from Motion Analysis, Inc. This thesis utilizes a data collection program developed at the laboratory and the data analysis software program from Motion Analysis, Inc. DATA ANALYSIS - DATA COLLECTION PROGRAM The computer program used for data collection was developed by Mr. Robert Wells, Michigan State University (1989). This program, named BELDATA, can be used for collection of kinematic data either separately or simultaneously with the collection of ground reaction force data or EMG (muscle activity) data. With the pilot studies conducted on the above knee amputee and the Olympic racewalker, an AMTI force plate was used for the simultaneous collection of motion and force data. For this research, BELDATA was used for motion data collection only. Motion data was triggered manually by the researcher and duration of collection was set for six seconds at a rate of 60 Hertz. For all free swing trials of the above knee bench prothesis, collection started prior to the leg being released from its starting position. BELDATA, for each 50 51 trial, created a separate video data file for each of the four cameras. These binary files are later converted to a format used for analysis with the ExpertVision software. AN OVERVIEW OF THREE DIMENSIONAL TRACKING The VP320 video processor accepts, synchronizes, and digitizes the video input received from the four cameras and stores this information on a SUN 4/260C workstation. The ExpertVision three dimensional software program is then used to initialize, track, and analyze this data. The tracking algorithm used by the ExpertVision system consists of two procedures - initialization and tracking. The initialization procedure associates video images with target names while the tracking procedure maintains target identities through space and time. The tracking algorithm uses the object coordinate data created through the calibration procedure as previously described in Chapter Three. In the calibration process calibration coefficients were established for each camera view. These coefficients together with image coordinates of a target defined a path of an optical ray through the object space. When calibration coefficients and image coordinates are available for two or more camera views, the three dimensional path of an optical ray is defined for each of the views (Motion Analysis, 1990). To initialize the data, a researcher must select a camera view and a frame of data within that camera view and 52 then identify a specific video image as a named target. It is best to fast forward to a frame of data in which all targets are visible. The process of identifying video images with specific target names is repeated until all images in that camera view have been identified correctly. For each pair of images, the ExpertVision software uses the calibration coefficients and direct linear transformation to project a pair of rays into the three dimensional space. With a least squares program, the system finds a point of intersection obtained from the projections of paired images in three dimensional space. The point was then tested to see if it represented a valid target location. To be accepted by the computer as a target, the point must be located within the space established in the tracking arguments and the norm of residuals resulting from the least squares calculations must be less than or equal to the value established by the researcher for the maximum norm of residuals (Motion Analysis, 1990). These tracking arguments or parameters are set by the researcher before the tracking process begins. Parameters include: a numerical description of the object space, the number of pixels that can be contained in a target, the maximum norm of residuals, the speed at which a target can travel, and the number of frames that a target can be absent from view before being considered lost. If the location is accepted, the target is assumed to be viewed by at least two cameras and the resulting three dimensional location is stored in a time- 53 space array. If the location is not accepted, the initialization process must be repeated using a different frame of data. After initialization, three dimensional target locations are defined for one frame of data. By direct linear transformation, these locations are projected into the two dimensional image space of the next frame of data from each of the four cameras and used to identify the two dimensional images of that frame. Once targets are identified in two or more frames, the predicted three dimensional locations, determined by linear extrapolation, are projected into the next frame and used to identify two dimensional images in that frame. Just as in the initialization process, the projected three dimensional locations for each target must be located within the object space defined and must not exceed the norm of residuals, both of which are set in the tracking arguments. Along with these two criteria, the distance between the previous three dimensional location of the target and its new three dimensional location must be less than or equal to the value established by the researcher for the instantaneous speed. If the location is accepted, the paired images is assumed to belong to the target. If the location is not accepted, the target is considered lost and will be dropped from the view until at least two cameras can see it. 54 TRACKING The tracking procedure allowed the researcher to define theoretical rigid links to approximate the motion of the above knee bench prosthetic as it underwent free swing types of tests. When the six targets were identified, links between the targets were established. Although not necessary for tracking, these links were visually beneficial to the researcher. Links were formed on the computer A- between the right and left sides of the hip axis, the right and left sides of the knee axis, the left hip axis and the left knee axis, the right hip axis and the right knee axis, p the left knee axis and the heel, the right knee axis and the heel, and the heel and the toe. Figure 14 graphically depicts this linkage system. The ExpertVision software then tracked the data for the predetermined time of six seconds. Once the tracking was complete, the data points were edited and smoothed, if necessary, using a digital filter on the track editor program within the software. The track edit program displayed the trajectory paths for the three dimensional data for each of the six targets tracked. It is at this point in the data analysis that a researcher can determine if a target had been reassigned incorrectly somewhere in the tracking process. If this is the case, the trial needs to be re-tracked from where the confusion occurred in the system. It is at this point, also, that a researcher can determine if the data makes sense. The track edit program was used for this thesis to eliminate any 55 ‘F—flrh Hip axis E Knee axis {Heel (Toe Figure 14: Linkage system for bench prosthesis 56 trajectories not associated with a named target, such as those formed from reflections, and to interpolate over any gaps in the trajectory paths. Once edited, the data was ready to be analyzed. DATA ANALYSIS - DATA ANALYZING PROGRAM The computer program used for data analysis was part of the ExpertVision three dimensional software package. The "angle" operator is a specific program which quantifies angular relationships between two line segments, each defined by a pair of objects and also a single line segment and a specified axis of the frame of reference (Motion Analysis, 1990). The angle operator required a three dimensional path file containing the trajectories of named targets as input. This path file contained the points which defined line segments in three dimensional space. Each line segment was defined by the locations of two named targets and both targets must have been tracked on a given frame for the line segment to be defined during that frame. The algorithm utilized for the angle program calculates the scalar, or dot product of two vectors. The angle operator calculates angles as the inverse cosine of the dot product of two vectors divided by the product of their magnitude. S7 ANGLE CALCULATIONS Angle calculations were performed on the data from the above knee bench prosthetic free swing tests used for this thesis. Employing the angle operator described above, knee flexion and extension angles were calculated using the first feature of the program in which angular relationships were quantified between two line segments each defined by a pair of targets. Hip flexion and extension angles were calculated using the second feature of the software. Toe movement was also calculated. The angular relationship was quantified between a single line segment and a specified axis of the frame of reference, the x-axis in this case. RESULTS AND CONCLUSIONS It was discussed in Chapter One that the swing phase of walking is controlled by the hamstring and quadricep muscle groups. The above knee amputee, without these muscles, loses swing phase control during gait. Prosthetic designs incorporated resistance mechanisms, or damping, at the knee joint to try to better control the prosthetic leg during the swing phase of gait. Resistance units reduced the angular momentum of the shank and prevented abrupt impact and subsequent rebound at full knee extension. Some damping also retards initial knee flexion during swing. In attempting to control the prosthetic swing phase, it was also discussed that efforts might be better directed at decreasing the periodicity of the shank by redistributing the mass of the shank. The above knee bench prosthesis used for this thesis did not have a damping system at the knee joint. Therefore, it was predicted that there would be impact and rebound after full extension of the knee. It was also predicted that changing the inertial properties of the prosthetic shank would enable the prosthesis to have greater knee flexion and a faster period at the knee joint. The rebound effect was seen in both the flexed start and straight start positions. Unfortunately, the accuracy of the free swing tests with the flexed start position was questionable because of a stability problem with the bench prosthesis. When the knee reached maximum extension, the 58 59 entire structure "bounced". This occurred despite efforts to stabilize the base of the structure to the floor with weights. Toe movement plots verified this observation (Figure 15) and because stability was a problem during testing, data from all bench prosthesis tests that had a flexed start position were dismissed. The bench prosthesis structure did not bounce during the tests with a straight start position. Data from the tests with a straight start position were reviewed and were found to be more characteristic of actual gait. Therefore, these results were used in the evaluation of the above knee prosthesis. The initial angle of the straight start position was inconsistent throughout testing. To adjust for this when reporting the results of free swing tests, all rebound angles and flexion and extension angles were converted to a percentage of the initial angle and then compared. No rebound effect or knee flexion was seen in the straight start conditions that had the knee extender rubber band attached. The surgical tubing used in the design of the knee extender for the above knee bench prosthesis was extremely tight and kept the knee in a locked position of extension throughout the tests conducted in that start position (See graph example Figure 16). For those test conditions in which the knee extender rubber band was left unattached, rebound effects were observed (Refer to Appendix A - Figures 17 through 28). The 6O rum AK BM "008“.“ v f ~ eta-easel.“ Deploy leap: ‘4' ‘t 0 rem lone-x 1 Lute-rem x ”it 0 v OMITIGI: an e» p 1 lb. wt 25 cm below knee Joint m’ 0 Fleaned start position “1» . J. ”1' 4h ”1? o A ”(P 4) S ”tr 1. o g “1’ 1» TOE a 1.9 m N no 0 “1" 0 12h -xe-ae-xe-n-e e dououoaaeobammmmononm Ycoord‘erdeCcr-n) Figure 15: Toe movement plot for flexed start position of bench prosthesis verifying "bouncing" of structure "to: flatware-men. Total Sort-o: 1 Oteptey W: rem Influx 1 Last 50H”: 1 "lot unt'nflhtp.“ Total Ice-tux I Dteplay longer rem who: 1 Loot lee-toe: 1 NE (301 1d) 2.312153-» HIP (open) meat" 61 - ENCH ' V I V ' V 7' C II x ' “T” No weight “ ‘““ ‘ Rubber band connected “ J‘ " * Strut t . star ; pox tion 68 . 0300 424- 41.» our .5» n ...,,VVJI .. ’ . -12 . d.» .24L -3. b 1b 15 I» an .‘3 .4... . .. v54 A . l - A 1 A 40 ‘ : ‘ ‘ - : ‘ ‘ 3 o ; - : s 5.5 0.5 5.5 5.5 1.2 1.5 1.5 ‘24 2.4 2.7 3.5 5.3 5.5 3.5 4.2 4.5 4.5 5.1 5.4 5.7 Time (second-z) Figure 16: Example of test conditions with knee extender attached and no weight 62 greatest amount of rebound effect was seen in the condition with no weight on the shank (Refer to Appendix A - Figure 17 and 18). The rebound effect was approximately 29 to 30 percent of the start position. The least amount of rebound effect was seen in two conditions. Rebound angles were approximately 20 percent of the start position with the one pound weight located 25 centimeters below the knee joint (Refer to Appendix A - Figures 23 and 24) and 19 percent of the start position with the 1.5 pound weight located 10 centimeters below the knee joint of the prosthesis (Refer to Appendix A - Figure 26). Knee flexion was also observed in the test conditions that did not have the knee extender attached. The greatest amount of knee flexion were seen in three separate conditions. The first, approximately 62 percent of the start position, was observed in the condition with no weight on the shank (Refer to Appendix A - Figures 17 and 18). In the remaining conditions with a weight attached to the shank, results were between 45 and 50 percent of the start position. The second greatest amount, approximately 46 percent of the start position, was observed in the conditions with either the one pound or 1.5 pound weight attached to the shank 10 centimeters below the knee joint of the bench prosthesis (Refer to Appendix A - Figures 21, 22, and 26). Initial observations indicated that there was a rebound effect with the above knee bench prosthesis, as predicted, 63 and that this effect decreased with a heavier weight on the shank attached close to the knee. It was also observed that knee flexion seemed to increase with the weight attached close to the knee. These results led this researcher to believe that redistributing the mass of the shank proximal to the knee may positively alter some of the parameters necessary to permit an amputee to have rapid gait. Studies conducted by Hicks et a1. (1985) verified that the period of the shank was .20 seconds faster with a mass attached 18.7 centimeters from the knee as opposed to 31.7 centimeters from the knee. Tests for this thesis showed that the period of the knee joint ranged from approximately 1.6 to 1.8 seconds. The shortest period of 1.6 seconds, was seen in the straight start position with the 1.5 pound weight located 10 centimeters below the knee joint. The longest period of 1.8 seconds, was seen under the same start position and weight condition, except the weight was located 25 centimeters below the knee joint. These results coincided with the results reported by Hicks and colleagues (1985), that the period was faster with the mass located proximally. In order to further investigate the effects of changing inertial properties of the prosthetic shank, cross plots of knee angle versus hip angle of the above knee bench prosthesis were created (Refer to Appendix B — Figures 29 through 40). To get good descriptive data for these cross plots, the six seconds of data collected for each trial was 64 "chopped" using a computer program written by Mr. James Patton, a graduate research assistant at Michigan State University. This computer program allows a researcher to shorten data files or "chop" parts of data files for more thorough evaluations of specific sections. The data files were shortened to include only the first two seconds of data so that concentration was on the first full free swing cycle of the bench prosthesis. The graphs show when the hip reaches a neutral position during free swing. With the knee extender rubber band unattached, the hip angle reached neutral at 0 to 27 degrees of knee flexion. Neutral, or 0 degrees, on these graphs was defined as maximum extension of the bench prosthetic hip. If the prosthetic knee reached maximum flexion at the same time the prosthetic hip reached maximum extension, the foot clearance would be maximized and the effectiveness of changing inertial properties of the shank could be evaluated. Ideally, in rapid gait such as running, maximum knee flexion occurs as the hip reaches maximum extension. For this research, the best condition occurred when the 1.5 pound weight was located 10 centimeters below the knee joint of the above knee bench prosthesis (Refer to Appendix B - Figure 38). Under this condition, the hip reached neutral at approximately 21 degrees of knee flexion with maximum knee flexion being 23 degrees at approximately 4 to 5 degrees of hip extension. 65 There is one parameter that is impossible to measure when evaluating an above knee prosthesis to permit rapid gait and that is the psychological response and motivation of an above knee amputee that may wear the prosthesis. No one except the amputee is able to "feel" differences and alter gait in accordance. Implications can be drawn when the test results from the above knee bench prosthesis are compared and combined with the results from the pilot studies conducted on the above knee amputee triathlete. Altering inertial properties of the bench prosthesis decreased the rebound effect, increased maximum knee flexion, and increased the speed at which the prosthesis swings. The best condition was found with a 1.5 pound weight placed 10 centimeters below the knee joint of the bench prosthesis. This same type of trend was seen during the above knee amputee's racewalk training sessions. In the condition with no weight attached to the prosthesis, knee flexion ranged between 60 and 62 degrees (Refer to Appendix C - Figure 41-43). Gait speed under this condition was 1.5 meters/second. With a 1.5 pound weight attached just below the knee of the amputee's prosthesis, knee flexion of the prosthesis was approximately 66 to 68 degrees (Refer to Appendix D - Figures 44-46) and gait speed was 2 meters/second. It was this condition, with a weight attached proximal on the shank portion of the prosthesis, that the amputee expressed confidence in feeling his leg "snap through" during swing phase. The swing time also 66 showed a slight decrease by .05 seconds. The above knee bench prosthesis proved to be a good model for an above knee amputee. Altering the inertial parameters on the bench prosthesis confirmed the results seen in the pilot studies with the amputee that a weight located proximal to the knee allowed for more rapid gait. For the above knee amputee, jogging normally with a foot-over—foot gait has always seemed to be impossible, but the results of this thesis indicate that given the right prosthesis design, rapid gait is possible for the amputee. RECOMMENDATIONS Recommendations for future studies include: 1 Altering the structural design of the bench prosthesis to assure stability during testing. The base support may benefit from a thicker and heavier material. 2 Using other types of rubber band knee extenders with various tensions. 3 Regulating the starting position for testing. 4 Creating a computer model to simulate an above knee prosthesis. 67 APPENDICES APPENDIX A FIGURES 17 through 28 68 FtIae strat'rtimaa.“ Iota) Sartaa: i - v - - W - 1 Diaplay Ian's: 455 ’ ., '33 3:21 1 «air common: Ma: atrat'etihtpan. Db Right 1"“ ”M" ‘ '7' ' Straight start position " thlay lama: L First Iartaa: i 4554) . we Sartaa: : one . «e ” ml , ME (solid) m’ _ flexion“) ‘10 I extension(-) e HIP (open) 1 fitensignh) ~in . aim -) -2.‘ e 40» 4° * .. ~50” 0 «it .. '"i’ 0 -ae . 4.0 0 once - - 1 g A A 1 - M I: ole eIe 1.2 :Is :.e 2:: 2.4 2:7 M :13 3:5 3.9 4:2 :sTe 5:: 5.4 5.7 Time (seconds) Figure 17: Knee and hip flexion and extension angles plot condition — no weight Ptlax ate-aim.“ Iota! 5cm: 1 ntaplay lama: Fir-at Iartaa: 1 Last 5artaa: 1 File: ate-“mu.” Iota) 3artaa: i Diqlay lanai First ”in: 1 Last. Iartaa: i KNEE (sol id) flexion (4») extension (-) HIP (open) assent" 69 AKBW‘I v f v ”0- common: in «lo - No weight .. m » Straight start position .. ~70 c..(> I) -i00 : : ‘ : ; ‘ : : ‘ r c c : : : : : ‘ 0.0 0.3 0.5 0.5 1.2 1.5 1.5 2.1 2.4 2.7 3.0 3.3 3.5 3.0 4.2 4.5 4.5 5.1 5.4 5.? Tara (seconds) Figure 18: Knee and hip flexion and extension angles plot condition - no weight 7O me: mm.“ AK BBC-l IataISartaari -4 i- m-.. - 0 I : ifi’wmf'... . no common. .. M “m" A «an 1 lb. 'gt at center of gravity y, m w . ,3“ ”MT" f“ m Straight start position 0taptay We rem arm: a on .. Laatlartaa: e 0..” o «e » . .’.ib KNEE (solid) e20- i (I . gratfload ’“’ HIP (open) ° "F—e-r extension“) 40* flexionI-I my o”t[ 401» ~50o .uj. . ‘701’ 0 “r o -ae~ .. -:eo ~ A A 4 A - A 0.. 2.2 0.. 5.. 1.2 1.5 1.. 2.1 2.4 2.7 2.. 3.3 3.5 3.2 4.2 4.5 4.. 5.1 5.4 5.7 Time (seconds) Figure 19: Knee and hip flexion and extension angles plot condition - 1 pound at center of gravity 71 "tee W.“ L J - AK BM # - . . 100a) Oar-teat i ' * f ' fl ' I 0t Ia hangar sir-3'»... e *0" common 1’ Laatlartaat i 0000 1 lb. egt at center of gravity-1 ";:L:m"r' nah Straight start position .L thlay w: rm: Iartaa: i ‘9. ‘T “stubs: i out .. a) “.1 p 930 KNEE (8012” .2. . Millet-) :10 ’ HIP (open) 01 eartension (e) -xe‘ f leerion (-) -2e . -ae J 4. -50<- -ee » v -7. r o 0.. h 4’ 40¢ 1* . a . A A A x A A L A A . - . A . A 1 0.. 0.3 5.6 5.. 1.2 1.5 1.. 2.1 2.4 2.7 2.. 2.3 3.3 3.8 4.2 4.5 4.. 5.1 5.4 5.7. Time (seam) d Figure 20: Knee and hip flexion and extension angles plot condition - 1 pound at center of gravity 72 "Ia: ate-mam.“ Ietallartaa: t am 51% ~ I flux“; ‘ m coenrrron: M 3"“= i m 1 lb. wgt 10 on below knee Joint Filer atrqtaihtpdn. Straight start position IataIIae-taat : m .. 0taptay lama: First screen: i 000” ‘* were... i we no 0,04» ME (solid) - flexion“) extension(-) .ee. HIP (open) . resent” -2. 4 '30 40 400 o.e e.x M M 1.2 1.5 x.e 2.: 2.4 2.1 a.e 3.3 3.5 3.1: 4.2 4.5 4.x 5.: are C1 Tune (accords) Figure 21: Knee and hip flexion and extension angles plot condition - 1 pound 10 cm below knee joint 73 Ft): W. 151a: Series: ‘ang : . - : -fi : OAK ng - - A. c—rm o;::l:’,.,,..: i one > COWIIION: 0 M “fl”: ‘ a. _ 1 lb. wgt 10 on below knee Joint j rqugmtpiw mm Straight start position 0 Display langax j Fir-at 5artaa: 1 0500 y. ' ‘ Laattartaa: i " 050+» .. 440. .80 b KNEE (solid) .20 - fiction“) extensiont-I “0' o HIP (open) 01L fitensiqnh) 4 action —) -1, , ’30“ «0 ~ 601+ 40» -70 r .. -ao L .. 0.00 sh -roo - - 1 0.0 0.3 0.5 0.0 1.2 1.5 1.0 2.1 2.4 2.7 3.0 3.3 3.5 3.9 4.2 4.5 4.0 5.1 5.4 5.7 Time (seconds) Figure 22: Knee and hip flexion and extension angles plot condition - 1 pound 10 cm below knee joint 74 feta: Minna.“ AK BEND! Fatal Series: 1 ’ v A v 4 - e e 'iw’wWIJ , m coemnom .. M “'1'" 1 “a 1 lb. wgt 25 cm below knee Joint j m an . neam‘ p‘llll a. Straight start position 0 Display hangar first. Iartaas 1 “'jr 4 " Laattartaa: 1 ’ m . «e 420 » KNEE (solid) our fiction“) . extension(-) no me (open) ° at (+) '1' ’ Hexionfl) -3. . -ggi «no -se » -se~ » '7. " U '00: e ‘04» 45 -:ee 4 ‘ ‘ o.e o.s o.s e.e 1.2 1.5 :.e 2.: 2.4 2.7 3.1: 3.3 3.s e.e 4.2 4.5 4Is s.: ate 517 Tune (seconds) Figure 23: Knee and hip flexion and extension angles plot condition - 1 pound 25 cm below knee joint 75 file: W.“ A 1 - AK am A _A You] wilt: 1 v v - a j—le- - fl 1 , f , ":3'3; , my, commas: “OH-"0" ‘ m” 1 lb. wgt 25 cu beluknee Joint “ me: We”... Straight start position rem Series: x 070‘- 00.)” hangs: Fir-at Oar-tea: i ‘49 Laattartas: 1 as «as net-z (solid) ‘” ‘ ‘ ) 42° ’ HEEHHR» “a HIP (open) m i ( ) “816622). -ee( -20 '7. P e» 400 0 40 > . r e.e e.s e.s e.e 1:2 1.: ale 2.: is 2.7 3.x 3.: a.s e.e 4:2 It 4.s st: 5:4 s17 Time (m) Figure 24: Knee and hip flexion and extension angles plot condition - 1 pound 25 cm below knee joint 76 Pile: may...“ ‘Iatai Series: 1 0iapiay lane: First Series: 1 Last. Series: 1 Filer Whip." Iota) Series: 1 Diqlay Ian's First Series: 1 Last. Series: 1 m (sol id) giggdgiédi-i HIP (open) 33.131 -1” - AK BENIN “" common: .. t“ 1.5 lb. wgt at center of gravity .. e10 » Straight start position A .A - e.e 0:3 e.s e.e 112 1.5 1.5 2.: 23¢ 2.7 3.5 3.3 3.5 3.5 4.2 415 4:3 5.: 5.4 5.7 1'ime (secortds) Figure 25: Knee and hip flexion and extension angles plot condition — 1.5 pound at center of gravity 77 rm: Mike's." A AK BDICH Iota) Iriaa: 1 - f - . - ~ 4 e - ”Numerous; i new (200111014: .L mum" ‘ 4.01» 1.5 lb. wgt 10 CI below knee Joint .. m. Mini . 10;, um" T" my, Straight start position 0 0iaplay Iaagax "m ”0.1 A “o I 1» Last Series: 1 KNEE (sol id) meme. 5:).-. HIP (open) anneal” 0.0 D 1» al.‘ h o -1 4 ‘ - : 4 ‘ e 4 : c - ‘ : : r ‘ c e 0.0 0.2 0.5 0.0 1.2 1.5 1.0 2.1 2.4 2.7 3.0 3.3 3.5 3.0 4.2 4.5 4.0 5.1 5.4 5.7 Time (seconds) Figure 26: Knee and hip flexion and extension angles plot condition — 1.5 pound 10 cm below knee joint 78 filo: amounts.“ tots! Series: 1 Oiqlsy W: First Series: 1 Lost Series: 1 F110: Minimum Iota! term: 1 Dispisy lungs: First term: 1 Lost lot-ice: i KNEE. (solid) flexion (4») extension (-) HIP (men) mam” AKBENGI «ow CONDITION: .. .go . 1.5 lb. wgt 25 c- belo' knee Joint .7. , Straight start position “0‘ 080 “I09 4 40L '00 > ‘ 400 fi‘ : ; : : ‘ ‘ : : : ‘ : f ‘ : ‘ : 0.0 0.3 0.5 0.0 1.2 1.5 1.0 2.1 2.4 2.7 3.0 3.3 3.5 3.0 4.2 4.5 4.5 5.1 5.4 5.7 Time (seconds) Figure 27: Knee and hip flexion and extension angles plot condition — 1.5 pound 25 cm below knee joint 79 Fiie: sit-Wang Vote) Series: 1 Display Range: First Series: 1 Lest Series: 1 File: emu.“ Total Series: 1 Display Range: First series: 1 Last Oeriee: 1 ME (solid) flexion“) sion(—) HIP (open) ( ) gamer % v r C—‘ : AK- BENL f v . fl . f ”0‘ CONDIIlON: ‘ m 1.5 lb. wgt 25 on below knee Joint Straight start position -70 b i» .00 > 0 40¢ A 4 A A A A c 0.0 0:: 0.5 m 1:2 C5 1:8 2.: 2.4 2.7 3.0 3.: 3:6 3T9 4.2 «TS (a s: 5.4 572 Time (secorc‘s) Figure 28: Knee and hip flexion and extension angles plot condition — 1.5 pound 25 cm below knee joint APPENDIX B FIGURES 29 through 40 80 av unwwoz o: I coauwvcoo oncm aw: m> meco mocx mo uoaq wmouu "om muswfim m-_oz< maze. . an ON a. o 0—- _ 35-.-... _ _ 8..- co. ON- On. 0?. 00. ON- a p- o— O? on an on cow 319M! le 81 av unwfioz o: I coauwwcoo mamcm aw: w> mawcm mocx mo uqu mmopo “om moswflm w-.02< wwz: . on em o— o op. _ _ _ In ’0 0 0 0 0 fiOfiv I! o 1| o/ I o/o/j/fl/ .1 i - \ z-m-uo - oil-Kiel.- I oi- ‘1 o .- !. DP.\\.\° ‘iiiil AVG. 0\ TI II.- e\~/\o\ \& .l 9/ 9o I. f/ L oil/o VII [Illi- /¢/¢I_ T 33.8 L i 00’. ca. on. 00. cm. 01. cu. o—- 9— cc cc on cop 3'!an le 82 cm muw>muw mo noucoo um canon a I :Ofiu oawcm a“: m> oawcm omcx mo uoHa flccoo mmouo "am ouswwm m..02< wwzx . 8 S o 2- _ . _ sooisllswmw... .- fi/o 4W6 .l \\ Klilllli Illilli: i: till-olilbli 33-3- 1 ox/w..\fi% .- 1 j-fa/0 .l. fil. ./0 l1 Ilia!!! @111. 1 .I III?!- dung I. rll L J co..- on- on- em. O7 00. o..- op ON. 0* on 00 an 8— S'ISNV le 83 ow >uw>mum mo poucoo um canon H I cofluficcoo ofiwcm afi; m> oncm mocx mo uoaa mmOpu "mm musmwm w402< wwzv. on an o. a o.- _ I I.. _ Ii $0-20 oII-io I. ¢// .1 ?/é/ ,I lé/do .l DIIIIIIIII-IlII-il0\.| \Q -II \-o Il}i-r¢\i- i- -Io\\ .M. .- 91- ,. M. .- ] Auk/b -| .l e// I oiIIIIo I. Ito-{Ii/o .I. IAYI L1 EWSNV le 84 unwoh oocx zoaon Eu OH venom H I :Ofluflwcoo oncm aw: m> oawcm mmcx mo uofia mmoou "mm ouswwm m-_Gz< mwzv. . ow on am e. c o—- _ _ . _ l.- fli oiIIIo-IIIIOIII-TR8OOI-L. iI .- M TI» ¢ I 4/././ a I av l -I II-.-©.I\o-III-II||.I-III0\IIII\-\ ..o...oII-\-I\ II oo.\ J - mi 4 I. IJPoilxo .l I //@/Q\Q 1 II /¢/¢/0i’/0 I1 i! o ABIXiRGQU fl .1 .l .L _ _ _ 8 T oo- oo- as. oo- cm- ov- on. ON- 0—- o— 8. on O? on on on 8’ 31st am" 85 on ucfloh omcx 30am: Eu OH venom H I :owuflccoo oncm can m> mamas mocx mo uqu wmouo "cm ounwflm m-_Oz< mmzv. an S c 2. -. _- 1 .- o/o l /o/ l o I ll- i-IIr'llIll‘i-li \VIII- I W-II I - ohm . -I-I- o-I-I /¢/’f’! l 0/0- '1 f/lIo/ II OIIIII’! 1 I o /Io/¢/¢- go oop- cop 319M! le 86 peach mocx 3oHon so mm ncsoq H I cofiufivcoo oncm ma: m> mecm omcx mo uOHa mmoou "mm mpamflm MJGZ mecm mmcx mo uoHa mmouo "om ocstm m-_02< mwzv. on em o— o.- _ 7 T. 1 II In .I i .I IL .I ll I i _ _ oop- oa- cm- on. on. ov- ON- 0—. o— 8. ow on an ac. 319W le 88 muH>mew mo emueoo um venom m.H I eOHqueoo onew mH: m> onem mmex mo uon mmoeo "mm weeme Boz< mmz: . on cm A: o 3- _ _ II @\o.\o-\ J - fi/fi - .l ¢//$f/? I. \.V \\\b o\o\-\I\Io oI3\\\b6 TI 9\ ll Q\ l // I. on. 0?. on. em- co 319NV:M4 89 on . I queoo v woe 30Hon 50 OH venom m H eOH uefio. onem mHe m> onem ooex no uon mmoeo m462< mwzz ON a— O “mm meeme OH- om- ov- an. ON- Op- 9— ON on ow on snow dIHT 90 on ueHOH ooex onon EU mm veeom m.H I eoHqueoo onem mas m> onem moex mo uon mmoeo Boz< mwzv. _ on c— 9 "mm oneme OT om- cv- on. ON- 2. av 319M! dIH 91 ueHOm moex onon 50 mm venom m.H I eoHqueoo onem mHe m> onem ooex Ho uon mwono "on onanm m-_OZ< mwzx on on o— c 3. _ _ .IIOO-WO .. All - L-y\\\\ V -..o \o\ o-Am-I I %\\o I-II-I: I. [- /o/ .1 -///o T /.0/ II o/o/ _ _ fa”. om. ov- cm. 0—- op. on O? on BTSNVcflH APPENDIX C FIGURES 41 through 43 92 Fiie: ensue... um arm: 1 I - —H '32:...“ . .- ; . i u“ 3"“: 3 . LIFT It WEE m mm ”115 um TRIAL } F 1e no... ‘ '14.? 3...?" 1 m mm (left solid) mmnw. (right open am) .- 01.)” w First kiss: 1 Lest Irina 1 7' ’ - s > ‘3 u > . ' i ‘ i “ > F “ h 40 ’ 1 . > 13 u L J» an I . I IL “ F I. t 1 s I ‘ , 1 .. 1.. 2.. 2.. 2.. 1.. 1.. 8.. 0.. 0.. 3.. 3.. 4.. 4.. t. C..4.. 5.. 5.. 5.. Tune (second) Figure 41: Above knee amputee knee flexion angle plot condition — no weight 93 FiieI dim... Yet-oi bin: 1 sup», by: First bin: 1 int win: 1 mammmxmummnxn " . nosnmnc (left solid) MIMIC“. (right open ate) - I I - 3 r J i - . ' - - I T - I - - ', i I 2.:- 23 2.. 2.13 2.. us 3.. 3.. 3.13 3.. 4.13 4.: «In Tune (sseold) Fits: “.33. Fetal ”z 1 Noisy up: First sci-z 1 test bin: 1 a v . v sell 0 Figure 42: Above knee amputee knee flexion angle plot condition - no weight 94 )ete) bin: 1 I a Why Ills-x ’ 33:31: "umumnnmmnauummuumcwmunnn ‘ ..-.,.......,.., u- nosmmc (left solid) “1.1011. (right open now - 7.“ '0.) I n ’ U Dicky Ile- ”M “'1‘: I J Lest series: 1 10 I . r H . P J» g . . ’ i) G r ‘ ‘ ’ i . P i . ’ I - i a ' ‘ . I _ I ‘ ’ ) II ’ I t ‘ .3" 9 ‘ t. i I b { ii in 1.. 8.10 1.. 2.. 2.1s 2.. 3.:s 3.9 3.. 3.1s 3.. 4.10 3.. 3.. Tune (eseold) Figure 43: Above knee amputee knee flexion angle plot condition - no weight APPENDIX D FIGURES 44 through 46 95 Fi)e: “til-u.” Vets) trio: 1 high, ”: First Orin: us: ”0.: Vi : LEFTAKWEEMFLDOM new mm: TRIAL I ‘ numeric (left solid) mm mm opu- am) I L - I g - P , ’ i - I I - - . ! - - I 3 - P l { ts 2.43 2.. 2.. 3.. 3.: 3.4a 3.0 3.. 4.. 4.33 4.43 4.. 4.. Tune (second) Fiie: mm.” Vets) 1.3.: 1 big)” up: First bin: 1 Lat kiss: 1 v ‘. Figure 44: Above knee amputee knee flexion angle plot condition - 1.5 pound below knee joint 96 '31s: ”11133.... 1.31 has: 1 01.1” up: first 333333: 1 Lset hm: 1 "13: m... tessl 333333: 1 01.137 .0: first 1.333: 1 un 3:333: 1 I “ALu!marmannansmnunmmumwnrun ‘ 33 L msnenc (left when mmnou. mun opu dots) J 7|» . aL . Ir a . “ b . ’ H ‘ > H ‘L 3r ‘r 33 ' . [ I k a v I 1 I 3 » o :3 ’ I . l ( . 1.1» 1 4 I} Y Q i. 3.33 3.. 3.33 3.43 3.. 3.. 4.. 4.33 4.43 4.33 4.. 3.. 3.. Tins (ssceml) Figure 45: Above knee amputee knee flexion angle plot condition - 1.5 pound below knee joint 97 "be: “13131.... 1.31 ”3.: 1 ' D‘splsy w: fix33§§ '“ mnaxmwnzmurummnumnuuwntmn . D ";'* WW mmnc (13:: solid) minim (up: open dots) m' h‘-’ ‘ a , n 0333‘” 33333: Hrs! 1.3-: 1 ‘ L331 1.133: 1 fl . r d . r ‘ U “ 3 1 . ’ , _ H ‘ F J» . b ‘ 4 ‘ b ' 1 I a » ,- a > ' 9 a > . ' 1’ r l 18 P 4 I, t :3 , . s i t I ‘ 0 =3.3 3.2 3.4 3.3 3.3 4.3 4.2 4.4 4.3 4.3 3.3 m 3.4 3.3 3.3 3.3 Turns (seam!) Figure 46: Above knee amputee knee flexion angle plot condition — 1.5 pound below knee joint L I ST OF REFERENCES LIST 0? REFERENCES 1. Andersen, R., "Running: A Road to Mental Health," Bunnsr_a_nerld V01 14.1979 PP 48- 51 2. Cairns, M.A., Burdett, R.G., Pisciotta, J.C., and Simon, S. R., "A Biomechanical Analysis of Racewalking Gait," Medicine and Science in Snnrts nng Exercise, Vol. 18, 1986, pp. 446-453. 3. Cappozzo, A. Pigura, F., Marshetti, M., and Pedotti, A., "The Interplay of Muscular and External Forces in Human Ambulation," lgnnnal_gi_fiinmeghanigg,Vol. 9,1976, pp. 35- 43. 4. Cappozzo, A., "Gait Analysis Methodology," Human_ Movement Science, 1984, pp. 27-50. 5. Clausen, J.P., "Effect of Physical Training on Cardiovascular Adjustments to Exercise in Man," thsielegical_zexieus. Vol. 57. 1977. pp. 779-815. 6. DiAngelo, D.J., Winter, D.A., Ghista, D.N., and Newcombe, W. R. "Performance Assessment of the Terry Fox Jogging Prosthesis for Above- Knee Amputees," Jgn;nnl_g£ Binmngnnnign, Vol. 22, 1989, pp. 543- 558. 7. Drillis, R., "Objective Recording and Biomechanics of Pathological Gait," Sciences, Vol. 74, 1958, pp. 86—109. 8. Enoka, R. M. ,Miller, D. I., and Burgess, E.M., "Below- Knee Amputee Running Gait " Amarisan_lgurnal_21_£hzsical Medicin , Vol. 61, 1982, pp. 67-84. 9. Farge, E.J., Hartung, G.H., and Borland, C.M., "Runners and Meditators: A Comparison of Personality Profiles," Jgnrnn; of Pensonality Assessment, Vol. 43, 1976, pp. 501—503. 10. Godfrey, C.M., Jousse, A.T., Brett, R., and Butler, J.P., "A Comparison of Some Gait Characteristics with Six Knee Joints," rthotics and EEQELhBLiQS. Vol. 29, 1975, pp. 33-38. 98 99 11. Hicks, R., Tashman, S., Cary, J. M. ,Altman, R. F. , and Gage, J. R. "Swing Phase Control with Knee Friction in Juvenile Amputees," 12urnal_2£_Qrthenaedic_nsssarch Vol 3. 1985, pp. 198- 201. 12. Roy, M.G., Whiting, W.C., and Zernicke, R.F., "Stride Kinematics and Knee Joint Kinetics of Child Amputee Gait," Azgnivng 9f Physigai Mgdigine nng Rehabiiitnti n, Vol. 63, 1982, pp. 74-82. 13. Internat'onal Amateur Ath et'c o ' , Marshallarts Print Service: West Sussex England, 1984, p. 162. 14. Ismail, A.H. and Young, R.J., "Effect of Chronic Exercise on the Personality of Adults" in Ihe Marathon: Physiological, Medical, Epidemiological, and Psychoiggignl Studies, edited by P. Milvy. Annals of the ugw ank Agnggmy of Sciences: New York, 1977, pp. 958-969. 15. James, U. and Oberg, K., "Prosthetic Gait Pattern in Unilateral Above-Knee Amputees," Scandinavian ignznnl g: Rehabilitative Medicine, vol. 5, 1973, pp. 35-50. 16. Kegal, B., Carpenter, M.L., and Burgess, E.M., "Functional Capabilities of Lower Extremity Amputees," Archives of thsical Medicine and Rehabilitation, Vol. 59, 1978. pp. 109-120. 17. Kegal, B., Webster, J.C., and Burgess, E.M., "Recreational Activities of Lower-Extremity Amputees: A Survey," Archives of Physical Medicine and Rehabilitation, Vol. 61, 1980, pp. 258-264. 18. Leavitt, L.A. and Zuniga, E.N., "Gait Analysis in a Series of Normal Subjects as Related to Above Knee Amputees," Isr. Journal of Medicine, Vol. 9, 1973, pp. 592-598. 19. Lewis, E.A., "Fluid Controlled Knee Mechanism Clinical Considerations," ulletin of Prosthetic Researcn, Vol. 10, 1965, pp. 24-56. 20. McCullough, N.C. III, Hampton, P.L., and Harris, A.R., "Prosthetic Knee Units" in Atlas of Limb Prosthetic; - Surgical and Prosthetic Principles edited by American Academy of Orthopaedic Surgeons, C.V. Mosby: St. Louis, 1981, pp. 386-387. 21. Moore W.B. and Malone, J.M., ang: Ennnenity Amputation. W.B. Saunders Company: Philadelphia, 1989, pp. 295-316. 100 22. Morgan, W.P. and Pollock, M.L., "Psychologic Characterization of the Elite Distance Runner" in Inn garathon: Physiological, Medical, Enidemioiogigni, and Psychological Studies edited by P. Milvy. Annalg Qt Lne Ne! York Academy of Sciences: New York, 1977, pp. 382-403. 23. Motion Analysis Instruction Manual, 1990. 24. Murphy, E.F., "The Swing Phase of Walking with Above Knee Prosthesis," Bulletin of Prosthetic Research, Vol. 10, 1964, pp. 5-39. 25. Murray, M.P., Sepic, S.B., Gardner, G.M., and Mollinger, L.A., "Gait patterns of Above Knee Amputees Using Constant Friction Knee Components," Bulletin of Prosthetic Research, Vol. 17, 1980, pp. 35-45. 26. Murray, M.P., Mollinger, L.A., Sepic, S.B., Gardner, G.M., and Linder, M.T., "Gait Patterns in Above Knee Amputee Patients: Hydraulic Swing Control vs Constant Friction Knee Components," Archives of Physical Medicine ang Rehabiiitation, Vol. 64, 1983, pp. 339-345. 27. Patil, K.M. and Chakraborty, J.K., "Analysis of a New Polycentric Above-Knee Prosthesis with Pneumatic Swing Phase Control," Journal of Biomechanics, Vol. 24, 1991, pp. 223—233. 28. Radcliffe, C.W. and Lamoreux, L., "UC-BL Pneumatic Swing Control Unit for Above Knee Prosthesis," Bulieiin g1 Prosthetic Research, Vol. 10, 1968, pp. 73-89. 29. Radcliffe, C. W. and Lamoreux, L.W., "The University of California Four-Bar Linkage Knee for the Above Knee Amputee" in Advances in External Control of Human Engneniiigg, edited by Gavrilovic and Wilson. Proceedings from the 4th International Symposium on External Control of Human Extremities, 1972, pp. 786-793. 30. Scheuer, J. and Tipton, C.M., “Cardiovascular Adaptations to Physical Training," Annual Revigg g: Physiology, Vol. 39, 1977, pp. 221—252. 31. Seyle, H., fitness Withou; Qistr98§, J.B. Lippincott: Philadelphia, 1974. 32. Smith, A., "A Biomechanical Analysis of Amputee Athlete Gait," International Journal of Snort Biomechanics, Vol. 6, 1990, pp. 262-282. 101 33. Stokosa, J.J., "Prosthetics for Lower Limb Amputees," in Vascular Surgery Principles and Technignes edited by H. Haimovici, M.D. Appleton-Century-Crofts: Norwalk, CT, 1984, pp. 1143-1161. 34. Wagner, E.M. and Catranis, J.G., "New Developments in Lower Extremity Prosthesis" in Human Linps and Their Substitutes edited by Klopstey and Wilson. McGraw-Hill: New York, 1954, chapter 17. 35. ‘Wallach, J. and Saibel, E., "Control Mechanism Performance Criteria for an Above Knee Leg Prosthesis," gournal of Biomechanics, Vol. 3, 1970, pp. 87-97. 36. Zarrugh M.Y. and Radcliffe, C.W., "Simulation of Swing Phase Dynamics in Above Knee Prostheses," Journal pi Biomechanics, Vol. 9, 1976, pp. 283-292.