The Walker Team Product Design Proposal February 9, 2006 Course: BME 227L Instructor: Aura Gimm Ian Shakil Katie Sullivan Mark Connell Chris Schumann We have neither given nor received improper assistance in the completion of this assignment. I understand that a violation of this assignment can result in failure of this course, and/or suspension from Duke University -2- Abstract ........................................................................................................................................... 2 Background ..................................................................................................................................... 2 Goal Statement & Functional Specifications.................................................................................. 5 Proposed Work................................................................................................................................ 6 Design Evaluation........................................................................................................................... 8 Conclusions & Next Steps ............................................................................................................ 10 Bibliography ................................................................................................................................. 11 Abstract Assistive walking devices are used by individuals who suffer from decreased mobility. However, only a few models exist to support a wide range of disabilities, from balance difficulties to leg amputations. Our clients, Dr. Dan Erb and Dr. Kathy Shipp from the Physical Therapy department at the Duke Medical Center, plan to redesign existing walkers in order to better serve their osteoporosis patients. Before doing so, they require data from the patient- handle interaction from existing walkers. The clients have requested a device that can measure the force that a user exerts on the left and right handle-grips of a walker or rollator during normal use. These force measurements will be measured in the Krzyzewski Human Performance Research Laboratory (K-Lab) over a 1-minute testing period. The resulting data will be used to prescribe walkers and rollators that minimize excessive loading in the user’s upper extremities. Ultimately, it is hoped that the data will be used to design superior walkers that maximize comfort, aesthetics, and load distribution. This device must meet the following criteria: • Measure the force exerted by each hand (right and left) onto the hand grips of a traditional and rollator walker • Not alter the user’s gait pattern significantly by adding height to the grips or weight to the frame • Output data that is easily synchronized with the K-Lab motion analysis system • Be portable and adaptable to multiple walkers • Be Easily set up and calibrated • Be easily cleaned and very durable The team has considered a host of design proposals ranging from floor-based motion sensor mats to instrumented walkers, and it was decided that a “portable sleeve – cantilever handle” (PSCH) device is the most feasible and cost effective design that satisfactorily addresses the needs of our clients. Background Osteoporosis, which translates into “porous bones,” is a disease in which bones lose calcium structure and become more likely to break. Over 200 million people worldwide are -3- affected, 80% of whom are women, and millions more are at risk.1 While onset is most common after age 50, osteoporosis can develop at any age. Throughout a lifetime, old bone is removed from the skeleton in a process called resorption, and new bone is added in a process called formation. Until around age 30, bone formation exceeds resorption. After peak bone mass is reached, resorption occurs at a more rapid rate than formation. In a subset of people, this process continues until the bone reaches a dangerously low mass (Figure 1). If left untreated, the structural deterioration of osteoporosis leads to bone brittleness and an accompanying increase in fractures, particularly in the hip, spine, and wrist.2 These fractures are serious; hip fractures frequently involve hospitalization and surgery, and spinal fractures can result in height loss, back pain, and deformity. Both can cause permanent disabilities and impair a person’s ability to walk unassisted for the rest of his or her life. Figure 1. A normal bone (right) and an osteoporotic bone (left). The osteoporotic bone has lost a significant amount of structure, which leads to frailty and a higher likelihood of fractures3. Patients who experience decreased mobility as an effect of osteoporosis are frequently prescribed an assistive walking device. Assistive walkers and rollators are also prescribed to patients suffering from post-surgery trauma, balance problems, and general frailty. These devices effectively allow the user’s upper extremities to provide added support and stability for the patient. The type of walker prescribed is patient specific and depends on the extent of required balance and weight-bearing assistance. Canes provide the least support, followed by crutches and walkers. Assistive walkers are intended to increase mobility, prevent falls and fractures, and encourage independence. The two most common walkers are the standard aluminum walker and the rolling walker, or rollator (Figure 2). The standard walkers are constructed from a tubular aluminum frame, plastic handgrips, and rubber-tipped, height adjustable legs. At about seven pounds, they are relatively lightweight, however the patient must have the upper body strength to lift the device and move it ahead after each step. Additional drawbacks include a slow and awkward gait and the possibility of a curved, kyphotic spine. Use of this walker is usually limited to indoor, single storied buildings. The rollator is better suited to patients who lack the upper body strength and balance to lift the walker with each step. It is composed of metal tubing and has wheels on the legs to promote movement. Rollators often include a built-in seat in which the user can rest, a waist high safety bar to prevent falls, and a front basket for convenience. Patients with unsteady arms and hands may find the rollator unstable as the wheels react to any applied forces and could move in an unintended direction. In addition, the device is cumbersome and awkward, does not -4- allow the user reach forward, and cannot easily be maneuvered upstairs. Again, it is primarily designed for indoor use. Figure 2. Existing assistive walking devices. Conventional aluminum walker (left) and rollator (right). While both devices provide stability, they have drawbacks as well and neither was specifically designed with the osteoporotic patient in mind4. We are working with two doctors from the Physical Therapy Division of the Duke Medical Center who plan to design a new assistive walking device for osteoporosis patients to confront the drawbacks of the standard and rolling walkers. Dr. Dan Erb is an associate clinical professor in the Department of Community and Family Medicine. His past research has focused on the functional outcomes of physical therapy interventions in both the elderly and neurologically impaired patients. Dr. Kathy Shipp is an assistant professor at the Duke Center for Aging. She has researched osteoporosis and the resulting vertebral fractures, spine deformity, and rehabilitation. Before designing a novel walker, the clients have requested quantification of the various forces applied by the user’s upper extremities on the walker, both compressive (down) and sheer (forward and back). These measurements will determine whether the existing walker configurations cause harm to the joints, as well as to ensure that their future design will operate properly under these forces. The force sensors must not alter the patients’ normal pattern of ambulation through additional weight or height. The clients also stressed portability and adaptability of the force sensor to a number of existing devices. Previously, instrumented walkers have been reported in scientific literature. These devices use strain gauges to measure forces from the users’ upper extremities. In each instance, the walker has been permanently altered; none of the devices offer the portability requested by our clients. The group of Bachschmidt, et al, developed a strain gauge-based instrumentation system for six load measurements—three dimensions of forces and three moments.5 Twelve strain gauge bridges were placed within the handle. Simpson’s rule was used to compute stresses and strains through the cross-section, and Gaussian integration allowed for the calculation of moments and axial forces. The group of Pardo, et al, also measured walker reaction forces using strain gauges placed on the legs of a standard walker.6 The forces from each of the four legs were summed to -5- yield the total walker ground reaction forces. Walker stability was assessed using the calculated horizontal position of the center-of-pressure. The results were also used to identify upper extremity strength requirements. The group of Fast, et al, placed their strain gauges on the walker legs as well, at 15 inches above ground.7 Axial and bending forces on the aluminum frame were measured in both frontal and sagittal axes. The interface box was not attached to the walker in order to avoid extra weight; instead, the examiner carried it during each trial. Software and calibration data were used to convert the millivolt output into pounds. G. Simoneau of Marquette University recognized that walkers require weight bearing of the user’s upper extremities and adjustments in the kinematic pattern of both arms and legs.8 His instrumented standard walker, or dynamometer, measured the six loads on the handle using 24 double-pattern resistive foil strain gauges. A customized frame allowed static calibration, and cross-talk was corrected through data processing. The dynamometer trials were synchronized with a six camera Vicon motion analysis system. The results indicated significant joint moments at the shoulders, elbows, and wrists. Another group, Ming, et al, developed a dynamometer using twelve strain gauge bridges positioned around the walker according to Finite Element Analysis.9 The strain gauges were powered with a 12 volt DC source, and the system was statically calibrated using a multiaxis frame. Commercially available instrumented walkers are manufactured and sold by Advanced Mechanical Technology, Inc. (AMTI), a company in Massachusetts.10 Their devices examine the forces and moments involved with walkers, canes, and crutches, with such applications as medical research, orthopedics, and rehabilitation evaluation. Specific uses include gait and stability analysis. The instrumented walkers are available in three styles: pediatric anterior, pediatric posterior, and adult anterior. Strain gauge load cells output six measurements: three orthogonal forces and three moment components along the X, Y, and Z axes. The sensors have high stiffness and sensitivity, low cross-talk, excellent repeatability, and stability over many trials. Goal Statement & Functional Specifications The goal of this design project is to create a portable device that measures the forces applied to each of the grips of a walker or rollator. There are many functional specifications indicated by the client. First, the device must measure the compressive force applied to each walker handle over the course of a one-minute in-lab test. Ideally, all three dimensions of forces and all three moments would be measured. However, the compressive force is of the highest interest. Second, the device should be portable. The clients would like to be able to attach the device to any walker that a patient brings in order to characterize how that patient uses their own walker. Therefore, the device should not damage the walker, and should be easily attached and removed. The device should not alter the gait of the patient using the walker. Their use of the walker should remain as usual. The data from the device should be able to be synchronized with -6- the three dimensional motion analysis system found at the K-Lab. 1 A simple calibration scheme should be used, and the device should be durable and easily cleaned. Proposed Work Ultimately, the portable sleeve cantilever handle (PSCH), as shown in Figure 3, was selected as the proposed design. This design features a hand plate attached to a load cell to determine the force applied by the user. First, the load cell will be mounted in front of the current hand grip location. The mounting technique will differ depending on the load cell selected for the device. Attached through the threaded load cell will be a rigid support that links the cantilever handle to the mounted cell. This will create a large moment on the load cell but should not affect the performance of the device. A rigid support between the two key components is essential to the efficacy of this design, and different materials will be studied to optimize the performance. Figure 3. Cantilever Handle Design Sketch and Potential Mounting Mechanism. The top figure describes the cantilevered hand support, while the right figure illustrates the potential bracket and platform mounting system. Four other alternative designs were examined. The first design was implementation of a force pad. This would allow for the measurement of three dimensions of forces and moments at a single point on the pad, and would not modify the gait of the user. This design has a few drawbacks. First, the force pad measurements are only valid if there is one contact point; with a walker or rollator and a user, there are six contact points. A net compressive force could be determined, but much information would be lost. The size of a force plate would limit its usefulness. The largest plate produced by AMTI is 1.2m x 1.2m, which would not be large enough for a user walking with a walker for one minute. Also, while net vertical forces on the walker would be measured, the client is most interested in upper extremity forces exerted specifically onto the hand-grips. 1 According to Dr. Queen, Coordinator of the Human Performance Research Laboratory (K-Lab), it is only necessary for the proposed walker device to output a small analog voltage difference. The K-Lab equipment is capable of providing further signal conditioning and amplification. -7- Walker leg boots were also considered. A boot would be placed on each walker leg and would be able to capture the compressive force applied to the walker by a user. These boots sre composed of a cylindrical walker leg casing with a load cell at the bottom of the boot, surrounded by a rubber sole. It would slip on to a walker leg as shown in Figure 4. The boot would be secured by a rubber ring near the top. The leg boots would add no more than 0.75’’ to the length of each leg; the extra height could be compensated for by adjusting the leg length. This would have some effect on the user’s gait. The wires from each load cell would be attached to the corresponding leg and then bundled and strung from the back of the walker. Figure 4. Walker Leg Boots Sketch. The boot is tubing that fits around each walker leg, adding a short length and measuring the ground-walker leg force interaction. An instrumented walker design would effectively characterize the user’s employment of the walker. Such a design would replicate the instrumented walker created by Bachschmidt et al, as illustrated in Figure 5. The handles would first be removed from the walkers, and twelve strain gauges would then be carefully applied to a pipe or tubing similar to a walker bar. Next, the two instrumented grips would be mounted in the laboratory walkers. This method offers limited portability between reconfigured walkers. It necessitates precise engineering and calibration, but would allow for the measurement of the six moments and forces of interest. Figure 5. The Instrumented Walker Handle. The handle is outfitted with twelve strain gauges in various orienations to capture the three forces and three moments applied to 11 the walker . -8- The portable embedded model, shown in Figure 6, would achieve the high degree of portability desired by the clients. An array of thin-film load cells would contact both the hand plate and walker bar, as shown. A potential array configuration (Figure 6) would be two by two; however, this array size could be optimized during design modification. All forces applied by the user upon the handle plate would then be applied to the walker through the thin-film load cells, and the net force in two dimensions could be determined computationally. This design would not be excessively bulky and would only require two clamps at each end of the device for stabilization on the walker. By understanding the geometry of the walker grip (ideally a cylinder) and the compression forces, the four points could be summed to create a net force vector (vertical and lateral axes). Figure 6. Portable Sleeve - Embedded Handle sketch. The sketches illusrtate how the hand plate would rest on the walker handle, at 4 contact points with with compressive load cells located at each of the contact points. Design Evaluation In order to choose between the different designs, a decision matrix was constructed using five different criteria, each of which was assigned a weighing factor based on their relative significance. The most important design criterion was feasibility—a measure of how likely design could be realistically completed given the current restraints (time, available technology etc.). Feasibility was assigned a particularly high weighing factor (25 out of 100). The second category that was used to compare the different designs was whether or not the device would significantly affect how a person normally moves with the walker—a primary client concern. Thus, because of the importance of this requirement, it was also assigned a high weighing factor (25 out of 100). The third design criterion was cost, as it is imperative that the expenses necessary to construct the design within the established budget. This category was assigned a relatively high weighing factor (20 out of 100). The next category used to judge the designs was portability. It is extremely important for the client to be able to transfer the device from one walker to another. This was not one of the essential requirements of the client, and as a result, it was a given a slightly lower weighing factor than the previous categories (15 out of 100). -9- The final criterion was performance, or how well the design would be able to measure the desired forces. This includes the ability of the device to measure force in more than one spatial direction, as well as the ability of the device to distinguish between the different forces with mimimal crosstalk. While considered a very important criterion, the performance of the device was assigned a slightly lower weighing factor than previous categories (15 out of 100). Using these five categories, the following decision matrix was constructed on a 1-to-10 scale. As can be seen in Table 1, the force pad design scored a relatively low 655 points (out of 1000). Table 1. Decision Matrix for Evaluating Designs Feasibility Does Not Cost Portability Performance Total Alter Gait Score Weighing Factor 25 25 20 15 15 Force Pad 5 10 2 10 6 655 Walker Leg Sensors 6 8 5 7 6 645 Instrumented Walker Frame 5 6 5 0 10 525 Portable Sleeve-Embedded 7 8 9 10 7 810 Sensors Portable Sleeve-Cantilever 9 8 9 10 9 890 Handle (PSCH) The primary strength of this design was that it had little effect on how a person moves with the walker, since no changes would be made to the walker itself. Also, in terms of “portability,” it would be ideal, since any walker could be used with the force pad. However, the design received mediocre scores in performance (since it would be difficult to record forces other than compression) and feasibility (since force pads tend to be much smaller than what would be desirable for motion with a walker). But the primary detriment to using a force pad was the cost, as force pads are simply too expensive to be seriously considered as a possible design. The walker leg sensor designed also received a fairly low score (645). This design had fairly mediocre scores across the board, as it did not stand out in any category. While the design is feasible, it is somewhat less practical than the other designs just because it would require four devices (one for each leg) to be built while the handle grip designs require only two. The leg sensors received a relatively high score in the “does not alter gait” category, since as long as the sensors do not add too much weight to the walker, it is unlikely that the devices will have much of an effect on how someone moves with the walker. The mediocre scores that the leg sensor design received in cost and portability can again be attributed to the fact that four separate devices for force are required, making this design more expensive and more difficult to transport from walker to walker. Finally, the design received a relatively low score in performance simply because it is unlikely that any forces will be able to be measured with this device other than compression. It should be noted that the walker leg sensors could not function with wheel-based rollators. The design involving instrumenting the frame of the walker received the lowest score (525). This design received a perfect score in performance, since it is possible to position strain gauges throughout the walker’s frame to measure force in all directions. However, in the other - 10 - areas, this design scored mediocre to very low. Instrumenting the frame of the walker would be a difficult process requiring very precise engineering, thus resulting in a fairly low feasibility score. Also, all of the changes to the walker frame could have an impact on how a person normally moves with the walker, and as a result, a mediocre score was given in the “Does Not Alter Gait” category. Furthermore, according to AMTI research, instrumenting a walker would be expensive, likely exceeding the budget limitations. Finally, this design scored a zero in terms of portability, as the changes made to the instrumented walker would not be transferable to other walkers. The design involving a portable sleeve with the embedded sensor received a total score of 810. This device scored well in almost all of the categories. It received its lowest scores in feasibility (due to the difficulty of embedding very small sensors into the hand grips), and performance (since the embedded sensors would only be able to measure compression). The design scored well in “Does Not Alter Gait,” as it is unlikely that a slight change to the current handle grips will affect a user’s motion with a walker. Furthermore, the device scored an almost perfect score in cost, as small sensors have been found that are very inexpensive. Finally, the design received a perfect score in portability, as the device will be able to be easily attached and detached from any walker. The final design, the portable sleeve with the cantilever handle (PSCH), scored the highest, receiving an 890 out of a maximum of 1000. The device received a high score in every category. First, the design seems very feasible, as placing the sleeve adjacent to the handle grips removes the size restrictions on the force sensors that made the embedded sensor design somewhat impractical. The device also scored high in the “Does Not Alter Gait” category, as though the design may result in the hand grip being slightly higher than in a conventional walker, it is unlikely that such a small change will have any affect on how a person normally moves with the walker. Furthermore, from the research conducted, it appears that the design can be completed at a relatively low cost. However, the cost is dependent on the quality of the force sensor, which in this design, could range from a simple strain gauge to a complex three- dimensional load cell. The device thus scored very high in performance, as it would not be difficult (though it may be too expensive) to implement a three-dimensional load cell into this design, and thus be able to measure all of the desired forces. Finally, the design received a perfect score in portability, as it will be able to be easily transported from one walker to another. The portable sleeve with the cantilever handle is the design of choice. This design meets all of the necessary criteria: feasibility, affordability, portability, minimal gait affect, and effective performance capability. While the portable sleeve with the embedded sensors also scored highly and will likely be an excellent second option if the primary design fails, the portable sleeve with the cantilever handle is the best design for accomplishing the client’s requirements. Conclusions & Next Steps The client requires a device that can measure the force exerted on the left and right grips of a walker/rollator over a 1-minute time period within a clinical setting. This device must - 11 - synchronize with the K-Lab motion analysis system, be portable, cleanable, durable, and easily calibrated. It is also imperative that the device does not affect the normal gait of the test subject. The team has decided that the “portable sleeve – cantilever handle” (PSCH) device (Figure 3) is the most feasible and cost effective design that satisfactorily addresses the needs of our client. The PSCH is portable, universally compatible with all walkers/rollators, minimally invasive, durable, and cleanable. The PSCH design makes use of a miniaturized load cell rigidly attached to a cantilever- styled handle grip, which overlays onto the existing walker grip. The team must now begin an exhaustive search to identify a cost-effective load cell that provides an appropriate force- measurement range (0-80 lb)12 while withstanding large moments (~40 ft-lbs) and moderate heat fluctuations. The team must also begin designing, machining, and testing the rigid load-cell/grip connection, clamping mechanism, encasement, calibration system, and ergonomic cantilever grip. The first step in the process is to embark on a thorough materials selection search. Bibliography 1 Reginster, JY and Burlet, N. “Osteoporosis: A still increasing prevalence.” Bone, 2006 Jan 30;38:4-9. 2 National Institutes of Health. Osteoporosis and Related Bone Diseases National Resource Center. www.osteo.org. 2004. 3 Osteoporosis and Osteoporosis caused by Parathyroid Disease. Norman Endocrine Surgery Clinic, Tampa Bay, FL. www.parathyroid.com/osteoporosis.htm, 2006. 4 Your Health Fitness Store. Total Body Works, Weatherford, TX. http://www.healthfitnessstore.com/walkers/. 5 Bachschmidt, RA, et al. “Walker-Assisted Gait in Rehabilitation: A Study of Biomechanics and Instrumentation.” IEEE Transactions on Neural Systems and Rehabilitation Engineering. 2001 March; 9(1): 96-105. 6 Pardo, RD, et al. “Walker User Risk Index.” American Journal of Physical Medicine and Rehabilitation. 1993 Oct; 72(5): 301-5. 7 Fast, Avital, et al. “The Instrumented Walker: Usage Patterns and Forces.” Archives of Physical Medicine and Rehabilitation. 1995 May; 76: 484 – 491. 8 Simoneau G, Hambrook G, Bachschmidt R, Harris G. “Quantifying upper extremity efforts when using a walking frame.” Pediatric Gait:A new millenium in clinical care and motion analysis technology, Harris GF and Smith PA (editors). IEEE Press, Piscataway, NJ, 2000, p. 210-216. 9 Ming, Dong, et al. “A New Dynamometer Walker System for the Measurement of Handle Reaction Vector.” Measurement Science and Technology. 16(2005) 1272-1280. 10 MCW Walker Sensors Brochure. Advanced Mechanical Technology, Inc., Watertown, MA. www.amtiweb.com, 2005. - 12 - 11 Bachschmidt, RA, et al. “Walker-Assisted Gait in Rehabilitation: A Study of Biomechanics and Instrumentation.” IEEE Transactions on Neural Systems and Rehabilitation Engineering. 2001 March; 9(1): 96-105. 12 The client expects that, during peak loads, the walker-user will place a maximum of 80% bodyweight onto the handle grips. 80% of a larger 200lb patient would yield 160lb. Thus, each hand grip must be able to endure 80 lb (1/2 of 80% * 200).