I. Amputations and Limb Prostheses


A. General



Kevin Garvin, MD; Jerzy Novack; Donald Giger, PhD; Steven Hinrichs, MD
Omaha VA Medical Center, Omaha, NE 68105; Department of Pathology and Microbiology, University of Nebraska Medical Center, Omaha, NE 68198-1080

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #A2161-RA)

PURPOSE--Infection following total joint arthroplasty is a significant clinical problem, and its accurate detection is a matter of great value. The purpose of this study is to develop and evaluate a new detection system employing nucleic acid amplification (NAA) procedures. Previous investigations have found unacceptably high false positive results from NAA tests. To improve the sensitivity and specificity of NAA for application to musculoskeletal infection, a new approach to targeting bacteria has been developed.

METHODOLOGY--The primary assay for bacterial NAA employs the use of the polymerase chain reaction following the extraction of nucleic acids from submitted clinical specimens. To improve the sensitivity and specificity of techniques reported in the literature, a new approach was developed using primers targeted to selected species commonly encountered in periprosthetic joint infections, namely Streptococcus and Staphylococcus. Computer analysis of gene bank sequences identified appropriate primers for use, and these were synthesized and tested for application in a PCR assay. Samples tested in the first two phases of the project included those from a rat model of musculoskeletal infection and human samples, including tissue and joint fluid. Verification of effectiveness was completed using a rat model of musculoskeletal infection. Subsequently, 208 samples from 88 individuals undergoing total joint revision for suspected infection were studied, using the genus-directed approach. Samples were also submitted for culture by routine methods. Reviews of all clinical information were conducted on the 88 cases, and sedimentation rate, white blood count, and x-ray findings were included in the analysis.

PROGRESS--Within the first year, the validation of the assay was completed and effectiveness demonstrated using the rat model of musculoskeletal infection. The study showed the genus-specific primers were 100 percent in concordance with culture results from the samples obtained from infected rats. The genus-focused primers targeting Staphylococcus and Streptococcus species were then utilized in the evaluation of samples from humans suspected to have infected joints. The 16s ribosomal gene was the targeted sequence. A total of 208 specimens from 88 patients undergoing total joint revision were studied. Comparison was made with two reference methods, including standard bacteriologic culture and a second method of clinical impression in which all data were incorporated. Eighteen cases were positive by culture with bacterial species capable of being detected by the assay. In comparison to culture, the NAA genus-focused approach was 89 percent specific, having been positive in 16 of the 18 cases. However the PCR was positive in 41 of the 67 cases where no bacteria were detected by culture.

FUTURE PLANS--The next stage of this project is to expand the genus-focused primers to include primer sets targeting bacteria other than those in the Staphylococcus and Streptococcus groups. This is necessary to raise the overall level of sensitivity for clinical application. Additional efforts will be focused on improving specificity of the assay. The key issue remains as to whether the NAA approach is detecting true infection or contamination. In all 88 cases, a decision had been made to replace the joint prior to obtaining specimens for analysis, and it is unknown to what degree undetected or low-level bacterial infection may have contributed to the joint failure. Additional studies will seek to determine whether specimens for NAA are becoming contaminated at some point following surgical removal and prior to amplification.



Adrian A. Polliack, PhD; Samuel Landsberger, ScD; Donald R. McNeal, PhD; Dana Craig, BS; Robert Sieh; Edmund Ayyappa, MS, CPO
Rancho Rehabilitation Engineering Program, Rancho Los Amigos Medical Center, Downey, CA 90242; PACT Gait Lab, VA Medical Center, Long Beach, CA; email: polliack@ranchorep.org

Sponsor: None listed

PURPOSE--Socket fabrication is a highly refined art that relies on the skill and experience of the prosthetist. Despite best efforts, patients often return to the prosthetist with complaints of residual limb pain, socket discomfort, and skin problems. Clinical techniques for the quantification of biomechanical factors at the socket interface are not in common use. The concept of measuring pressure at the interface between the residual limb and the prosthetic socket could provide valuable information in the process of prosthetic socket fabrication, modification, and fit. The Rincoe Socket Fitting System (Rincoe & Associates, Golden, CO) and F-Socket Pressure Measurement System (Tekscan Inc., South Boston, MA) are two commercial systems designed for in-situ measurement of interface pressure. Their use is not common in prosthetic practice, perhaps due to cost, difficulty in clinical interpretation of the data, and time required for operation. Of greater concern is the use of sensors for pressure measurements in areas of high contour and complex geometries such as the residual limb. Furthermore, these systems have not undergone any scientific evaluation for validity in prosthetic applications. In order to assess the sensors' performance and their clinical validity, a series of trials was conducted to evaluate different aspects of sensor performance; namely, accuracy, hysteresis, drift, and the effect of curvature.

METHODOLOGY--All tests were performed in either a flatbed chamber or a custom-built pressure vessel. Transducer outputs were recorded over a range of pressures: up to 83 kPa for the Rincoe system and 552 kPa for the F-Socket system. Accuracy was defined as the percentage difference between applied and observed pressures by incrementally increasing applied load. Drift tests were done up to 20 min. All trials were conducted at ambient temperatures of 23.5-25 °C. Calibration of the F-Socket transducers were performed according to the manufacturer's specifications prior to each trial. The Rincoe sensors were calibrated initially by the manufacturer. To assess the affects of curvature, a plaster positive mold of a transtibial residual limb was used. A total of nine sites were chosen for pressure measurements, representing pressure sensitive and tolerant sites typically experienced by persons with transtibial amputation. Pressure transducers were placed and secured using medical paper tape. A silicone ICEROSS® liner was rolled onto the mold to secure transducer placements and prevent air leakage between the pressure vessel and the mold. The mold was then inserted into a custom-built pressure vessel, attached to an air compressor and a regulator valve to control inlet pressure level.

  After extensive testing of at least 125, but fewer than 400, trials, results indicated an accuracy error for the Rincoe system of 29 percent, with a corresponding 4.4 percent error in hysteresis and 6.2 percent error in drift. F-Socket system, however, demonstrated an accuracy error of 8.2 percent, a 24 percent error in hysteresis, and 14.4 percent error in drift. Additionally, the effect of curvature, by use of the positive mold versus the flatbed chamber, generally increased the error response for all tests. Furthermore, there was an improved response with new sensors versus older sensors (>90 days). Calibrating the sensors prior to testing illustrated an important criterion for use, as it would appear to allow for more accurate results. Hence, the preliminary results indicate a relatively favorable outcome for the F-Socket System compared to the Rincoe System. However, it is the authors' belief that both systems should be used cautiously and only to obtain relative and comparative pressure values.

FUTURE PLANS--Future considerations include testing the effects of temperature and humidity on these two sensor types. The influence of shear forces on normal pressure measurements will also be investigated. Clinical methods of data interpretation will be the last phase of testing for utility of interface pressure measurements, although the authors feel that this should not begin unless a sufficient sensor system exists for clinical purposes. The authors will begin to explore alternative sensor technology for prosthetics such as capacitance sensors. Clinical tests with a select group of transtibial amputees will begin to assess static and dynamic pressures and inter/intra-rater reliability of sensor application and testing.



Willem Atsma, MScEE; Bernie Hudgins, PhD; Dennis F. Lovely, PhD
Institute of Biomedical Engineering, University of New Brunswick, Fredericton, NB CANADA E3B 5A3; email: lovely@unb.ca

Sponsor: University of New Brunswick Research Fund, Fredericton, NB CANADA E3B 5A3

PURPOSE--The research goal was to explore the origin of patterns in the myoelectrical signals.

METHODOLOGY--The electrical signals from muscles have been used to control prosthetic arm devices for a number of years, and practical implementations of these myoelectrically controlled prostheses still use the same principles as described in the first paper published on the subject in 1948.

  The relatively recent finding that myoelectric signals contain a pattern that can be used to discriminate between different types of upper arm movements opened the way for research and development toward a new and more versatile paradigm for the control of powered arm prostheses. In this work, the origin of patterns in the myoelectrical signals is explored.

PROGRESS--The myoelectrical signals from six subjects and for four different movement patterns of the upper arm have been recorded and attempts made to classify these signals with a finite impulse response neural network, with mixed success. The goal was to provide a continuous classification method, thereby reducing the time until classification.

RESULTS--The best average classification rate achieved was 71.4 percent on a test set. A comparison with the static classifier indicated that the relatively poor classification rates are likely to be due to the recorded data. The comparison also indicated that the new classification method is a viable alternative to the original static classifier, while providing the reduction in classification time.



Mark Savoie, BScEE; Dennis F. Lovely, PhD
Institute of Biomedical Engineering, University of New Brunswick, Fredericton, NB CANADA E3B 5A3; email: lovely@unb.ca

Sponsor: University of New Brunswick Research Fund, Fredericton, NB CANADA E3B 5A3

PURPOSE--The purpose of this work is to design and evaluate a myoelectric control system fabricated around the HC11 single-chip microcontroller.

METHODOLOGY--The field of myoelectric control has been around for the last 30 years. While many advances have been made in the area of signal processing for such control systems, these advanced systems have been limited to research prototypes. In the area of commercial myoelectric control systems, little has been changed in the last decade.

  With improvements in HCMOS technology, the low-power, high-performance, single-chip microcontroller unit is now a reality. These highly integrated circuits are available with on-board memory, analog-to-digital (A/D) converters, and serial communication ports. With this flexibility, the development of a complex and versatile myoelectric control system has become a reality.

  This work is targeted toward the design and evaluation of such a myoelectric control system fabricated around the HC11. This single-chip solution is able to emulate any one of the various commercial myoelectric control strategies. In addition, this new design strategy holds great promise for the future as it lends itself well to remote monitoring and evaluation.

PROGRESS--A prototype system has been developed, with a user interface running under the Windows operating system. The hardware is capable of processing two channels of myoelectric information, while also accepting inputs from other transducers such as FSR or switches. System parameters are stored on internal EEPROM, and the therapist can view current signals in real-time with the connection of a host PC.


B. Upper Limb: General



Dudley S. Childress, PhD; Richard F. ff. Weir, PhD; Craig W. Heckathorne, MS; Edward C. Grahn; Jack Uellendahl, CPO; Yiorgos Bertos
Northwestern University Prosthetics Research Laboratory 345 East Superior Street, Room 1441 Chicago, Illinois 60611; email: d-childress@nwu.edu; web: http://www.repoc.nwu.edu/

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #A2087-RA)

PURPOSE--There is a need for better control of upper-limb, externally powered prostheses. Prosthesis control schemes that employ the body's own sensing systems seem to be incorporated more readily by prosthesis users. Such schemes may result in more subconscious control than other control schemes. It has been shown that position control is superior to velocity control in positioning tasks. But the majority of these prostheses are still controlled using switch or open-loop velocity control. Furthermore, open-loop velocity control cannot provide the appropriate sensory feedback. We believe that extended physiological proprioception (EPP) controllers as applied to externally powered prostheses by Simpson is a way of achieving this subconscious control.

METHODOLOGY--To explore these ideas, an analog EPP controller was developed and fitted to a number of persons with upper limb loss. Experience gained with this controller, and the availability of low-power, low-cost micro-controllers, have led us to develop a microprocessor-based EPP controller. The primary reason for going to a microprocessor-based design is one of flexibility: such a system allows us to linearize the nonlinear characteristic of the force sensitive resistors (FSR) we use to transduce force into voltage. This need was exposed by the observation that the analog device did not operate well at low forces and excursions.

  In the control of an externally powered prosthesis, the controller and FSR form part of a force-actuated position servo-mechanism. The FSR, along with signal-conditioning electronics, transduces an applied force to a microcontroller-readable DC voltage. This microcontroller then linearizes relationship between input force and output drive voltage using a look-up table. The new linearized value is then output as a pulse width modulated (PWM) signal to an H-bridge that drives the DC motor. User parameters can be changed, via rotary dip switches or the appropriate analog potentiometers, to tailor the controller to a particular person.

PROGRESS--A working prototype has been developed, and we are currently in the process of creating an initial surface mount prototype for further evaluation. Prosthetic fittings are planned to take place in order to test the device clinically. The EPP controller is designed around the PIC16LC73A microcontroller, a chip specifically developed for embedded applications where small space, low power, and versatility are of paramount importance. It consists of a central processor unit (CPU), 5 A/D input channels, 2 PWM output channels, EPROM, RAM, and 15 general I/O channels--all integrated onto a single chip that sells for less than $6. The current prototype can operate at a clock rates as low as 32 kHz, yielding a total power consumption of 250 µA for the controller. We ultimately envision using this controller as a building block for a multiple degree-of-freedom prostheses.

RESULTS--This controller has a much more linear response than the previous analog one. In addition, its parameters can now be changed independently of each other, greatly improving overall function. Much of the functionality of the controller is implemented in software, giving us flexibility in configuring different topologies (i.e., uni- or bi-directional). It can also be tailored for the force variations of a particular person during the course of using the prosthesis.




Trilok N. Monga, MD; William H. Donovan, MD; Diane J. Atkins, OTR; Robert L. Abramczyk, ME
Houston VA Medical Center, Houston, Texas 77030; The Institute for Rehabilitation and Research (TIRR), Amputee Program, 1333 Moursund, Houston, Texas 77030; email: monga.trilok_nath@houston.va.gov

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #A855-RA)

PURPOSE--To date, no quality of care (QOC) standards have been developed to address the rehabilitative and ongoing support needs of persons with upper limb (UL) loss. The lack of such standards may be contributing to less than optimal outcomes for these persons. The purpose of this project was to develop and test QOC standards for this population in VA and non-VA settings.

METHODOLOGY--This project was designed to employ an expert approach, similar to the "national consensus" approach employed by the Commission on Accreditation of Rehabilitation Facilities (CARF), to develop and test QOC standards for such persons. An Expert Advisory Panel developed the Amputee Services Inventory (ASAI) that included standards and indicators, a protocol for self-assessment, and site-visit testing of the standards and indicators.

  The ASAI was tested at five VA (Houston, Palo Alto, Seattle, New Orleans, and Hines) and at five non-VA (The Institute for Rehabilitation and Research, Rehabilitation Institute of Chicago, University of Colorado, University of Michigan, and University of Virginia) facilities. The following parameters were studied: how patients were referred, patient evaluation, preprosthetic therapy, prosthetic prescription, prosthetic fabrication, prosthetic check-out, prosthetic training, and prosthetic follow-up. Additionally, prosthetic facility/clinical standards of performance were reviewed as well as outcomes measurement of UL prosthetic services.

  The amputee program physician, occupational therapist, prosthetist, nurse, and program coordinator were included in the assessment and interview process, as were persons with past and present IL amputation (in person and by telephone). Mailed questionnaires (in English and Spanish) were sent to the most current such clients from each hospital.

PROGRESS--Amputee Program staff at each location completed a facility survey and coordinated a follow-up site visit by one of the investigators. Patient surveys were mailed to 78 VA, and to 138 non-VA, patients (216 total). Only 2 of the 10 sites (both non-VA) were able to provide the names of nearly 50 former patients, while the rest provided significantly fewer. Only 11 VA and 29 non-VA surveys (40 total) were returned to the investigators. Patients who did not respond to the first mailing were sent the survey again approximately 2 mo later. In addition to the mailed survey patient list, each facility was asked to provide up to 5 patients for interviews, either in person or by telephone. Ultimately, 5 VA and 15 non-VA patients were interviewed.

RESULTS--There appears to be no significant difference in the management of UL amputation in VA and non-VA settings. However, because there are very few of these patients in the VA system, it is difficult to make a valid comparison between the two. The rehabilitation care of these patients appeared satisfactory in all settings. Because there are relatively few persons with UL amputation currently being served within the VA system, we recommend the establishment of a "Center of Excellence" for them. This would enable the clinical/prosthetic team to become well experienced and proficient in the rehabilitation treatment of this often complex patient.



Thomas Bianchi, DrEng; Daniela Zambarbieri, DrEng; Giorgio Beltrami, DrEng; Gennaro Verni, DrEng
Dipartimento di Informatica e Sistemistica Università di Pavia, Via Ferrata 1 27100 Pavia Italia; Centro Protesi INAIL, Via Rabuina 14 Vigorso di Budrio (BO) Italia; email: thomas@linus2.unipv.it; dani@unipv.it giorgio@bioing2.unipv.it

Sponsor: Centro Protesi INAIL via Rabuina, 14 Vigorso di Budrio - Bologna - Italia; email: cprotesi.budrio@inail.it

PURPOSE--The aim of this work was to demonstrate experimentally the possibility to develop a new optical sensor that works with light in the infrared region to control an upper limb prosthesis.

METHODOLOGY--Near Infrared Spectroscopy (NIRS) is a noninvasive technique that uses the differential absorption properties of hemoglobin to evaluate skeletal muscle oxygenation. This technique (NIRS) relies upon two important phenomena: a) biological tissue is relatively transparent to light, and b) in tissues there are compounds whose absorption of light is oxygenation status dependent. We investigate muscle contraction using a NIRS commercial unit and EMG commercial electrode simultaneously.

PROGRESS--In the first part of the study, different experimental protocols consisting of flexo-extensions of the hand on the forearm have been tested in a population of control subjects. The protocols have been defined in order to simulate on the muscle contractions made by persons with amputation to control the prosthetic device in different conditions.

  In the second part of the study, the output of the NIRS instrumentation has been interfaced with a standard prosthetic hand. The interface has been provided by a software, developed in LabView language, that performs signal acquisition and processing.

RESULTS--The NIRS signals measured during contraction of the forearm agonist muscle present the same shape of variation as the signals revealed from the skin by using the EMG electrodes normally implemented in a myoelectrical prosthesis. In particular the 810 nm signal seems to be the most significative one, since it provides information on blood flow modification which is the major visible effect during muscle contraction.

  By considering the intra- and inter-subject variability of this signal in the controls, it has been possible to demonstrate that the polarity of variation is always positive, in spite of differences in the peak value. The same experimental protocols have also been tested on persons with transradial amputation with the same results.

  When the NIRS instrument has been connected to the power unit of the prosthetic hand, the controls have been able to produce hand opening or hand closing, depending on the activated muscle.

  Finally, the system has been tested on three persons with transradial amputation who normally use a myoelectric prosthesis. These subjects also were able to control the hand grasping in the same way as they perform with the myoelectric prosthesis.

  The results so far obtained suggests that the use of NIRS signals could represent a new alternative method for the control of upper limb prosthesis that does not suffer from the problem of interference and perspiration that sometimes affects the EMG electrodes.

FUTURE PLANS--A specific optoelectronic sensor will be developed, able to replace EMG electrodes without major modifications of the prosthetic socket. It will take into account the mechanical and size requirements needed to fit the sensor inside a prosthetic device.




André A.M. Sol, BSc; Dick H. Plettenburg, MSc
WILMER group, Department of Mechanical Engineering, Delft University of Technology, Mekelweg 2, 2628 CD Delft, The Netherlands; email: a.a.m.sol@wbmt.tudelft.nl

Sponsor: Delft University of Technology; Julianalaan 134, 2628 BL Delft, The Netherlands

PURPOSE--The standard split hook prosthesis is, despite its functionality, most often rejected by parents of a child with an upper limb defect because of the very poor and deterring outward appearance. The objective of this project is to develop a new prosthetic prehensor for these children that combines the functionality of the standard split hook prosthesis with an improved and appealing outward appearance.

METHODOLOGY--Initially the development was strongly focussed on appearance. A shape study was performed to determine the outline of the new prehensor. The resulting outline is derived from the contor of a hand of an infant. The length of the fingertips and the position of the rotating finger are approximately similar to a healthy hand. The connection to the forearm is harmonic and smooth. All mechanical parts, including the operating cable, are placed out of sight into a frame. Integrated within the frame is a lightweight friction wrist prosthesis. Also the frame is the pillar to the cosmetic cover made out of a flexible polyurethane resin. In this way the outside of the prehensor is rugged and easy to maintain; the cover can be easily removed to access the mechanism; and the cover color can be chosen from a wide palette. Giving the cover a bright primary color emphasises the toy-like nature of the prehensor, thus advancing the acceptance and use of the prehensor by the child.

RESULTS--In collaboration with our clinical partners of the Netherlands rehabilitation centers De Hoogstraat and Sint Maartenskliniek, 10 children clinically tested the cosmetic prosthetic prehensor. They all appreciate their new device greatly. It has not caused any negative reactions or strange associations. The children are delighted by the bright colored appearance of the prehensor. Because of the smooth outline of the prehensor and the integration of the control cable wear of clothing is reduced considerably. Some shortcomings were revealed as well. The mechanism of the prehensor, although reliable, is rather cumbersome to assemble. The characteristic of this mechanism had to be adjusted, especially in the case of shoulder harness control. Moreover, the flexible colored covers broke down too quickly.

  Therefore a new generation has been designed and constructed. A four-bar linkage mechanism is applied to reduce the input forces. An inclining input characteristic ensures the controllability for both shoulder harness and elbow control. The output force is almost constant over the range of opening. This prehensor successfully passed a series of durability tests in the laboratory. Also a new material for the colored coverings has been found. It combines durability with sufficient flexibility to permit easy exchange of the cover by the parents of a child or the child itself.

FUTURE PLANS--Clinical testing of the new prehensor mechanism is expected shortly. Successful completion of these tests will result in commercialising the cosmetic prosthetic prehensor.




Richard Q. van der Linde, MSc
WILMER group, Department of Mechanical Engineering, Delft University of Technology, Mekelweg 2, 2628 CD Delft, The Netherlands; email: r.q.vanderlinde@wbmt.tudelft.nl

Sponsor: Delft University of Technology, Julianalaan 134, 2628 BL Delft, The Netherlands; University of Twente, P.O. Box 217, 7500 AE Enschede, The Netherlands; Technology foundation STW, P.O. Box 3021, 3502 GA Utrecht, The Netherlands

PURPOSE--Previous research has shown that passive (or ballistic) walking is an energetic efficient and mechanically cheap way of walking. Therefore ballistic walking would be suitable for applications in rehabilitation technology and autonomous robots. However, ballistic walking behavior is determined solely by intrinsic system parameters. Fixed system parameters imply lack of a possibility for adapting or changing the walking pattern, and of a possibility to intervene when disturbances must be controlled actively. This severely limits the practical use of ballistic walking.

  The purpose of this project is to develop methods of synthesis that make it possible to apply the ballistic walking principle in rehabilitation technology.

METHODOLOGY--Theory and practice support one another in order to form a firm basis of insight. Return map analysis gives insight in the influence of system parameters on the dynamics of the mechanical oscillating bipedal system, and stability can be quantified. This way, the open loop behavior of the system is optimised as much as possible.

  Further, feedback can be added to increase the orbital stability, or to adjust the mechanical oscillation. Therefore adjustable passive components are needed, currently realized by energetically optimised McKibben muscles.

PROGRESS--A two-dimensional model has been developed, and analysed on periodic behavior. The model parameters were chosen to be equivalent to human mass and geometry parameters. Reflex-like trigger signals activated the muscles for a short period of time. A stretch-reflex activates the push-off plantar flexor, and a kick-reflex activates a hip stance leg extensor and a contralateral flexor. Limit cycles could be constructed and stable walking motions were the result. Changing parameters such as joint stiffness resulted in a shift of a fixed point, changing the resulting walking motion.

PRELIMINARY RESULTS--An actively variable passive stiffness joint has been developed. It facilitates to combine the advantages of passive motion with active control. This joint has been implemented in the hips of a biped and the leg extension joint. Walking motions were observed. However, stable walking is yet to be accomplished. Simulations show that this might be caused by a small basin of attraction of a fixed point. Currently a controller that helps the system to a fixed point is being designed.




Thomas M. Kennedy, BEng; Stephen Naumann, PhD, PEng; William Cleghorn, PhD, PEng
The Institute of Biomedical Engineering and Mechanical Engineering Department, University of Toronto, and Bloorview MacMillan Centre, 350 Rumsey Road, Toronto, ON M4G 1R8, Canada email: naumann@utcc.utoronto.ca

Sponsor: Natural Sciences and Engineering Research Council

PURPOSE--This research project is involved with the development of a multiple function control system for upper limb prostheses for children and young adults. In the past, various parametric representations of the myoelectric (EMG) signal have been chosen as input to pattern recognition algorithms (classifiers), either for simplicity's sake, or because it was hypothesized that they would yield a better result. However, no comparison of their relative merits has been conducted. Furthermore, clinical implementation of the classifier has seldom been considered in depth. Thus, investigations have almost always been conducted in controlled situations only, which does not allow for the large variations in the EMG signal under normal conditions of use (i.e., loading, fatigue, residual limb movement). It is the purpose of this work to examine the EMG signal produced under various conditions of use, and analyze it in order to determine the optimum parametric representation for use as an input to a classifier.

METHODOLOGY--Test subjects with below-elbow amputation or congenital limb deficiencies were asked to perform six different types of contractions (the contractions will be different for each individual, depending on personal ability) under controlled conditions, loading of the residual limb, and during motion of the residual limb. The raw EMG signal was recorded from two electrode sites, corresponding to those used for their conventional prosthetic.

PROGRESS--Further subjects are still necessary for completion of the project, and data analysis is currently underway simultaneously. The data are represented using a selection of spatial and spectral parameters calculated over various length time windows. The statistical package (SPSS) was initially used to cluster the data, with the hope that data would cluster into distinguishable groups according to type of contraction. However, the nonlinear behavior of the EMG signal has been shown to be too great for linear statistical methods. Currently analysis using nonlinear methods (feed forward artificial neural networks) has been initiated in the attempt to achieve better results.

FUTURE PLANS--Data collection will be completed pending the availability of appropriate subjects, at which time data analysis will be completed.




Bernie Hudgins, PhD; K. Englehart; P.A. Parker, PhD; J. Hayden, CET; D. Lynch, CET
Institute of Biomedical Engineering, University of New Brunswick, Fredericton, NB E3B 5A3; email: hudgins@unb.ca

Sponsor: The Natural Sciences and Engineering Council of Canada (NSERC); Whitaker Foundation; Hugh Steeper Ltd.

PURPOSE --A transhumeral amputation or congenital limb deficiency removes many arm functions, including the ability to position the arm (elbow and wrist functions) and grasp objects (hand function).

  A prosthetic limb is often fit to the person with such an amputation for cosmetic reasons and to replace some of the lost function. Myoelectric control has been used for many years to control the function of an artificial limb. These systems are routinely fit to persons with transradial amputation to control the grasping function of an electric hand. However, the extension to the control of more than one device, so that more of the lost function can be replaced, has been difficult. For the individual with high-level amputation (transhumeral or higher), and especially for those with bilateral amputation, the need for improved control systems for multifunction prostheses is critical to allow the return to work and performance of activities of daily living.

METHODOLOGY--During the past 7 years the Institute of Biomedical Engineering at UNB has been collaborating with Hugh Steeper Limited (HSL) of Roehampton England in the design of a multifunction myoelectric control system that addresses this limitation. Over the course of several collaborative projects between our two groups, and partially funded by NSERC through the collaborative research and development program, we have developed an intelligent myoelectric control system. Based on current signal processing and computer technologies, this system learns to recognize the myoelectric signatures from contractions the user has chosen to select each of the prosthetic limb functions. These contractions represent an intuitive link between the desired limb function and an easily repeated natural residual limb contraction. The trained system is then used to select and control the elbow, wrist, and hand functions of the prosthesis by simply reproducing the associated contraction. The new control scheme improves upon current commercial technology by not only providing all-myoelectric control of the limb but also by allowing the user to control the limb in an intuitive nonfatiguing manner.

PROGRESS--The hardware for the new control system has gone through many design iterations. It now meets the criteria for battery power efficiency (6-8 hr operation) and physical size (6.5×3.5×2 cm) to allow it to fit within the shell of a prosthetic limb. A total of 10 units, 2 trainers, and the associated training software have been manufactured. This new small control unit has been tested by clients in a clinical setting at our limb-fitting centre in Fredericton. As well, one of our transhumeral clients has been fitted with the device controlling all three powered functions (hand, elbow, and wrist); he used the system at our clinic and, although his evaluation of the unit was incomplete, his initial reports in which we have had him compare this all-myoelectric controlled limb to his current limb, (myoelectric hand, mechanically switched elbow, and passive wrist rotation) were very positive.

FUTURE PLANS--The control system is now ready for a more thorough and controlled evaluation. We are eager to demonstrate that our control system is an important contribution to the prosthetic community.



James Hunt, MScEE; Bernie Hudgins, PhD; Dennis F. Lovely, PhD.
Institute of Biomedical Engineering, University of New Brunswick, Fredericton, NB CANADA E3B 5A3; email: lovely@unb.ca

Sponsor: University of New Brunswick Research Fund, Fredericton, NB CANADA E3B 5A3

PURPOSE--The purpose of this research is to develop a less complex system that can accommodate present technology.

METHODOLOGY--There has been considerable research in the development of control strategies and techniques for controlling multiple degree-of-freedom (DOF) prosthesis. One such strategy is DOF control using pattern recognition to classify the myoelectric signal (MES) to select a DOF (at the hand, wrist, or elbow). Once the DOF is selected, the speed of the device is then controlled.

  It has been shown that the initial phase of the MES contains information describable by a set of parameters. It is assumed that these parameters of the MES are repeatable for a given state of muscle activation. Pattern recognition extracts information, known as features, from the initial phase of the MES. These features are then used to classify evoked contractions of the user. This is the accomplished by comparing features of a muscle contraction to that of a feature template derived during user training.

  Currently researched pattern recognition techniques use numerous features and complex algorithms such as artificial neural networks (ANN) to classify a MES. Complex algorithms enable the classifier to determine a large number of patterns and furthermore select the corresponding DOFs. Presently, the most complex artificial arms, however, can implement up to a maximum of 3 DOF. Therefore, a less complex system can be developed to accommodate this technology.

PROGRESS--A simple pattern recognition scheme for controlling a 3-DOF prosthesis was investigated. This new scheme uses only the mean absolute value of the myoelectric signal to reduce the complexity of previous pattern recognition schemes. Results of the proposed system show that a simpler hardware system is possible without considerably reducing the classification accuracy.


C. Upper Limb: Transradial



Richard F.ff. Weir, PhD; Dudley S. Childress, PhD
Northwestern University, Prosthetics Research Laboratory, 345 East Superior Street, Room 1441, Chicago, IL 60611; email: rweir@nwu.edu; web: http://www.repoc.nwu.edu/

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #A2263-RA)

PURPOSE--In this new project we propose to develop an externally powered prosthetic hand for persons with transmetacarpal amputations that result in loss of all digits. In this case, all except the very simplest function of the hand is lost. However, the wrist is usually still functional and its motions, flexion-extension, radial-ulnar deviation, and supination-pronation are extremely valuable in positioning the hand in space. In the past there have not been prosthetic components functional enough or small enough to fit gracefully into the space available. Thus, the need for the device described in this proposal.

  Preservation of the wrist's motion maintains the ability to position the hand in space, this ability is critical to maintaining the dynamic cosmesis of a prosthesis. The eye is tuned to the unexpected. If a prosthesis is moved in a 'normal' way it will pass unnoticed by the casual observer. The wrist is essential to this normal motion. A partial hand prosthesis that is hand-like yet does not inhibit the wrist will have both dynamic and static cosmesis. By powering the fingers, a functional hand-like prosthesis can be achieved. The challenge in this instance is to be able to place the drive mechanisms, power source, and electronics in the highly confined volume that remains after accommodating the residual limb. The only space left is in the artificial digits themselves.

  A powered-finger hand does not have to be limited to the fitting of partial-hand amputations but may also be of use in the fitting of wrist disarticulations and long transradial cases, where current powered fittings are frequently too long. The design would permit the length of the hand to be customized in these particular cases.

  We propose to build upon our previous experience to develop a new artificial hand that will be capable of developing 53.38 N of pinch force, have a width of opening of 8.9 cm and travel at rates in excess of 2 radians/s. It is proposed that myoelectric signals from superficial intrinsic muscles of the hand will be used to control the device in partial hand fittings. This prosthesis will be applicable to prosthetic restoration of prehension (with cosmesis) in persons with partial hand, and wrist disarticulation amputations, as well as the more conventional transradial fitting. Being light in weight it will also have application in the fitting of persons with high level transhumeral amputations.

METHODOLOGY--The new hand will use multiple motors operating in synergy with each other to provide reasonable force and speed capability. In a synergetic motor system, one motor is dedicated to providing high torque at low speed while another is dedicated to providing high speed at low torque. Together it is possible to have a system that is capable of achieving reasonable torque and reasonable speeds.

  We envisage two alternative designs: the first is a three-motor system in which the thumb is fixed and only the fingers move. In this configuration there would be a motor in the index and middle fingers to provide force, and a third motor lying in the line of the knuckles to provide speed. The knuckle motor provides speed to the opening of the fingers. The second is a four-motor system in which both thumb and fingers move. In this configuration there would be a motor in the index and middle fingers to provide force, and a motor in the thumb, in addition to the one lying in the line of the knuckles to provide speed.

  Through the use of carbon-graphite mounting components and the availability of small DC motors, we believe that an aesthetically pleasing lightweight prosthetic hand with a pinch force of 53.38 N and a speed in excess of 2 radians/s is readily achievable.



Just L. Herder, MSc; Dick H. Plettenburg, MSc; Jan C Cool, MSc
WILMER group, Department of Mechanical Engineering, Delft University of Technology Mekelweg 2, 2628 CD Delft, The Netherlands; email: j.l.herder@wbmt.tudelft.nl

Sponsor: Delft University of Technology; Julianalaan 134, 2628 BL Delft, The Netherlands

PURPOSE--The objective of this project is the design of a voluntary closing hand prosthesis for persons with unilateral transradial amputation that affords them pleasant cosmetics, good wearing comfort, and easy operation.

METHODOLOGY--In voluntary closing devices, operating force corresponds to pinching. To fully make use of the feedback potential of this working principle, the prosthesis operating mechanisms should transfer the actual pinching force to the operating member as purely as possible. Therefore, friction should be absent, and the counteraction by the cosmetic covering should be compensated for. Moreover, the operating force in general use should be in the range of best sensitivity of the prosthesis-user interface. Among the practical problems to overcome, the opened resting position, which is generally not acceptable for cosmetic as well as practical reasons, seems to be the most important one.

PROGRESS--A concept has been developed that combines advantages of voluntary opening and voluntary closing. According to this concept, the prosthesis can be switched between a passive and an active state. In the passive state the hand remains closed, regardless of elbow movement or of shoulder movement in case of shoulder control. The prosthesis becomes active when an object is placed in the hand, independent of the position of the elbow. A coupling between operating cable and finger mechanism is established such that applying an operating force results in pinching. Releasing the pinching force and taking out the object deactivates the mechanism and the prosthesis returns into its passive state.

  A prototype is being designed. High efficiency Rolling Link Mechanisms are applied. The fingers of the hand mechanism employ an adaptive grasp onto an object to minimize the pinch force needed. Moreover, the frame of the hand mechanism is flexible in nature, mimicking the natural dexterity of the human hand.

FUTURE PLANS--A transfer function to match the normal pinching forces with operating forces in the human optimum sensitivity range will be validated by additional research. The prototype of the hand mechanism will be evaluated in a laboratory setting. Consequently, a clinical prototype will be designed and constructed.




Shechar Dworski, BSc; Stephen Naumann, PhD, PEng; Morris (Mickey) Milner, PhD, PEng
The Institute of Medical Science, University of Toronto; Rehabilitation Engineering Department, Bloorview MacMillan Centre, 350 Rumsey Road, Toronto, ON M2R 1Y8, Canada; email: naumann@utcc.utoronto.ca

Sponsor: Natural Sciences and Engineering Council of Canada

PURPOSE--Tactile sensors have been added to fingertips of a prosthetic hand to detect incipient slip of a grasped object. A closed-loop control scheme will increase the applied grip force until slip is no longer present. This is a supplement to the standard open-loop control scheme which allows the prosthetic user to set an estimated grasping force through the detection of EMG (electromyographic) signals. Under the new scheme, once an object is grasped, the slip-detection algorithm monitors and adjusts the state of the grip until a new user applied force is initiated.

PROGRESS--This project has reached a midpoint, where a control scheme has been designed and preliminary tests using a prosthetic hand have been performed (Warren D'Souza's MASc thesis, January 1996). The current setup consists of a VASI 59 hand equipped with sensors on the fingers. The sensors convert a mechanical vibration, such as the one created when an object slips across their surfaces, into detectable electric signals. The two signals are sent to a processor where they are filtered, amplified, and compared with each other. The rationale for the comparison is based on the observation that noise signals (bumping of the arm, for example), will be seen as identical signals by the two sensors, whereas the slip signal will not be the same at both sensors. Thus, comparison of the two signals allows for the similar parts (the noise) to be filtered out, leaving the slip signal behind.

  The prosthetic hand used in the preliminary study was equipped with a large 3:1 gear reduction motor, in order to more accurately control the grip force. All control was performed external to the hand, through a PC and development board setup. Under these laboratory conditions, the results of the slip-detection algorithm were very promising and indicate the potential benefit in a final application.

FUTURE PLANS--In order to be compatible with compliant fingertips, flexible sensors will be used instead of hard ceramic ones. These sensors will also detect pressure, in addition to the slip signal. These signals will be used to add more precision to the control algorithm, and possibly for giving direct feedback to the user. The small, high gear ratio used in the standard VASI 59 hand must be taken into account. This will lead to the investigation of the electrical input to grip force output produced by the hand, so that a more accurate output control signal can be determined. Finally, miniaturization of the control scheme must be achieved. Part of this will be done using the new processor board already made for the VASI hand.



Gerald Stark, CP, CPed; Maurice LeBlanc, MSME, CP
Hosmer Dorrance Corporation, Campbell, CA 95008; VA Heathcare System, (153), Palo Alto, CA 94304; email: gstark@usit.net; leblanc@roses.stanford.edu

Sponsor National Institute on Disability and Rehabilitation Research (NIDRR), U.S. Department of Education, Washington, DC 22202

PURPOSE--Persons with upper-limb amputation have a selection of prehensors from which to choose. There are body- and externally powered hands and hooks. The least popular of these options is the body-powered mechanical hand. They are generally hard, heavy, and have very low efficiency. About half of that inefficiency comes from the glove, against which the hand must work to open. These hands are a prime target for improvement. We simply do not have good body-powered hands to offer this population: some opt for externally powered hands, even though they may prefer a body-powered prosthesis.

METHODOLOGY--This project undertakes an anatomical approach to develop a low-cost, lightweight, soft, endoskeletal, mechanical hand that will be controlled by body-powered cable. Envisioned is an endoskeletal mechanism embedded in a self-skinning foam: a separate cosmetic glove is not required. The mechanism will consist of four active fingers and a pre-positionable thumb. The fingers will have multiple joints to offer natural-appearing motion and an adaptive grasp with more surface area of the fingers in contact with the object. The thumb will be positioned to allow a smaller, more precise grasp against the fingers or a larger, more powerful grasp against the palm. The foam will be a soft, self-skinning polyurethane formed over the endoskeletal mekchanism placed inside a hand mold. Pigments can be added and various skin colors achieved.

PRELIMINARY RESULTS--The new prosthetic hand features an endoskeletal mechanism embedded in a self-skinning polyurethane foam that provides realistic appearance and feel without the need for a separate cosmetic glove. It is voluntary closing and, therefore, offers variable prehension force. The metacarpalphalangeal and proximal and distal interphalangeal joints are active, giving full natural motion. The control cable is attached to the distal phalanx of each finger so that pull moves all four fingers. The thumb is passive but movable to allow opposition of the fingers against the thumb or the palm. Compared to current mechanical hands, weight has been reduced 50 percent, and the new hand requires less prehension force to hold objects, due to the increased surface area of the adaptive grasp. Because of its simplicity, the hand can be produced much less expensively than those currently in production.

  Besides developing a mechanical hand that is simple, light, and soft, a major objective was to improve the low work efficiency (20-30 percent) of present mechanical hands. It is difficult to measure the absolute work efficiency of the new hand. However, comparison was made of the new hand versus Otto Bock and APRL voluntary closing hands in the amount of prehension force required to grasp a series of objects. The new hand required 12 percent less grasp force than the Otto Bock hand and 24 percent less than the APRL hand to hold the objects tested. The new hand, as presently configured, requires only 1.2 cm of excursion. If arranged so that 2.4 cm of excursion were required, then the force requirement for operation would be halved due to mechanical advantage. Therefore, the force to hold objects would be 24-48 percent less that the Otto Bock or APRL hands tested and therefore less cable pull required by the user.




Samuel Landsberger; ScD, Julie Shaperman, MSPH, OTR; Vicente Vargas, BSME; Andrew Lin, BS; Richard Fite, CP; Yoshio Setoguchi, MD; Donald McNeal, PhD
Rehabilitation Engineering Program, Rancho Los Amigos Medical Center, Downey, CA 90242; email: alin@ranchorep.org

Sponsor: National Institute of Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 22202

PURPOSE--The project seeks to design and develop an improved body-powered hand for toddlers (1-4 years). Primary design goals are an affordable hand of acceptable cosmesis to provide useful grasp for children while requiring minimal energy input.

METHODOLOGY--After exploring several hand design options, the work has focused on an Easy-Feed concept. Designed with a metal endoskeleton fitted with compliant foam and a cosmetic covering, its shape approximates that of the hand of a two-year-old. The design offers toddlers acceptable cosmesis and grasp function without need of a cable. Energy for inserting objects comes from the child's opposing, sound hand. A self-energizing principle captures and constrains the object in the hand. As the toddler develops more strength and cognitive skill, the hand converts to cable actuation and functions in either voluntary opening or voluntary closing mode. Power augmentation systems assist children with weaker muscles to use an active cable and strong closing spring to achieve good grip with voluntary opening body power systems. The design principle is to keep concepts and components as simple and rugged as possible, and small-lot manufacturing costs low. A laboratory test of grasp performance was developed to provide a baseline measure of new hand designs' functional potential prior to evaluation in the clinic. The staff and patients at the Child Amputee Prosthetics Project at Los Angeles Unit of Shriners Hospital for Children are working with the Rancho design team on the hand's development and initial clinical testing.

PROGRESS--Development is proceeding on both the skeletal/mechanical element and the flesh/cover. The effort is to bring the hand to a point where the concept and major design ideas can be tested in the clinic.

  The mechanism: the hand is a metal endoskeleton with a thumb that opens to accommodate a 5 cm object. The distal joints of the fingers flex inward to allow the child to push in objects easily, but do not extend past 90°, thus constraining the object. Another degree of freedom approximates the carpal joints (near the wrist), enabling the hand to open passively beyond the 5 cm active opening and to flex with a wrist-like motion, giving better compliance in activities, such as handlebar gripping, that require wrist flexion and extension. Several prototype models that have been tested in the laboratory appear promising.

  The flesh/cover: the problem of creating a durable, flexible, low-cost filler and cover with good appearance is a long-standing challenge. Conventional vinyl gloves lack the stretch needed for full thumb opening, compromise mechanical efficiency, and stain quickly, while silicone durability is poor for the child. To address this challenge, investigators are developing new designs in consultation with experts in the fields of entertainment/animation, cosmetic prosthetic restorations, materials chemists, and glove manufacturers. Some promising developments suggest that an acceptable solution should emerge soon.

  Power augmentation systems: 1) a battery-powered amplification module works in conjunction with the cable, supplementing pull in a pre-set proportion using the same principle as power steering assists the operator of a car. This retains feedback through the cable and allows the child to use a stronger closing spring than would otherwise be possible. One model has been successfully tested; a smaller version is in development. 2) A force-amplifier module fits on the forearm like a wristwatch and exchanges some cable excursion for force is also under development. Since the child can feed objects into the hand without fully opening it, the reduced excursion should not restrict function.

FUTURE PLANS--Glove molding will begin soon. The hand endoskeleton is near to limited run production. Local clinical testing and a test of parental preferences for various prehensors began December 1998.


D. Lower Limb: General



Mohammad Dorostkar, MD; Edmond Ayyappa, MS, CPO; Jacquelin Perry, MD; Richard Chambers, MD; Judith M. Burnfield, PT; Lara A. Boyd, MPT; Ernest Bontrager, MS; Sreesha S. Rao, MS; Sara J. Mulroy, PhD, PT
Pathokinesiology Laboratory, Rancho Los Amigos Medical Center, Downey, CA 90242

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #A861-RA)

PURPOSE--Increasingly, amputation as a result of diabetic foot pathology is limited to the forefoot. Following a partial foot amputation (PFA) the distribution of body weight over the foot is altered, and may lead to ulceration of the residual foot. The lack of toes following toe (TA) and metatarsal amputation (MA) changes the normally contoured forefoot into an abrupt edge, and reduces the forefoot-lever during single limb support. Individuals with a ray resection (RA) have a narrowed base of support, with potential balance and force concentration problems. Delineation of the mechanics of walking following PFA may improve surgical, orthotic, and prosthetic management, and enhance the durability of the residual foot. Therefore, the purposes of our research are to: 1) compare the biomechanics of walking following PFA with normal; 2) quantify the pressure pattern of walking following various levels of PFA; 3) establish the impact of the different levels of PFA on the sound limb; and 4) determine the efficacy of orthotic and prosthetic intervention affects gait following PFA.

METHODOLOGY--Comprehensive gait analysis of both lower limbs is being performed for each subject, including: measurement of kinetic and kinematic data, electromyographic activity patterns in the residual limb and sound leg, foot pressure patterns, and isometric torque capability. All data are recorded with and without footwear.

PROGRESS--Forty-nine individuals with PFA have been tested (24 TA, 10 MA, 13 RA, 1 Lisfranc, and 1 Chopart). Data processing and analysis are ongoing.

PRELIMINARY RESULTS--One portion of our analysis has focused on the mechanics of the residual limb forefoot rocker. All PFA groups walked significantly slower than normal (TA=61 percent N, MA=57, RA=62). The values for ankle moment, ankle power, F2, and F2-rate all covaried with velocity, necessitating the use of adjusted means. The time of peak ankle dorsiflexion was significantly delayed relative to normal for all the PFA groups (TA=49 percent GC, MA=52, RA=51, NC=43). Terminal stance ankle moment and its timing were significantly different for the MA group relative to all others (TA=1.24 Nm/BWLL, MA=0.74, RA=1.15, NC=1.35). The peak value of ankle power absorption was significantly less for the MA group compared to the TA (TA=-0.66 watts/BWLL, MA=-0.46, RA=-0.56, NC=-0.55). Additionally, the time of the power absorption for the TA and MA groups was delayed relative to NC (TA43 percent GC, MA=44, RA=42, NC=37). The rise rate of vertical ground reaction force from mid- to terminal stance before reaching the F2 peak was significantly lower for the MA and RA groups when compared to normal (TA=0.15 N/percentGC, MA=0.12, RA=0.14, NC=0.20). However, the values for the F2 peak were not different (TA=10.7 N/BW, MA=10.2, RA=10.3, NC=10.8).

  Additionally we have investigated load forces on the sound limb. Peak vertical ground reaction forces were higher on the sound versus the residual limb for the MA group (11.3 and 10.3, respectively), but not for the RA and TA groups. Sound limb peak vertical ground reaction forces inversely correlated with residual foot length for both MA (r=-0.89, p=0.007) and RA (r=-0.808, p=0.028) groups. After adjusting for velocity, all PFA groups had significantly higher sound limb peak vertical ground reaction forces compared to NC (9.7 N/BW). MA adjusted force (11.9 N/BW) was significantly higher than for TA (11.0 N/BW). RA (11.3 N/BW) was not significantly different from either the MA or TA groups.

FUTURE PLANS--Data collection and analyses continue with a target number of 70 participants. Specifically, the weight-bearing pressure patterns of both feet are being investigated to determine the efficacy of prosthetic inserts and footwear in the preventing of ulceration. The impact of partial foot amputation on loading patterns of the sound limb, and unloading patterns of the residual limb is also being analyzed.




Steven A. Gard, PhD; Dudley S. Childress, PhD
Northwestern University, Prosthetics Research Laboratory, 345 East Superior Street, Room 1441, Chicago, Illinois 60611; email: sgard@nwu.edu; d-childress@nwu.edu; web: http://www.repoc.nwu.edu/

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #A2133-RA)

PURPOSE--We performed mechanical analyses of three commercially available vertical shock pylons (VSP) to gain a better understanding of their function during walking. VSPs are designed to absorb shock in lower-limb prostheses by attenuating forces associated with walking and high-impact activities such as running and descending curbs and stairs.

METHODOLOGY--We investigated the Flex Foot Re-Flex Vertical Shock Pylon, the Ohio Willow Wood Stratus Impact Reducing Pylon, and the Seattle AirStance Pylon. Static and dynamic testing of these devices was performed using a mechanical testing apparatus. Static testing involved slowly loading and unloading the pylons while measuring both the force applied to the pylon and the pylon's linear displacement. The dynamic testing involved step loading and unloading the pylons while measuring the resulting axial displacement.

  The Re-Flex VSP consists of graphite telescoping tubes in parallel with a carbon fiber compression spring for shock absorption and energy return. There are a total of nine spring stiffnesses available from Flex Foot, accommodating subjects weighing anywhere from 100 to 325 lbs. We tested five of the nine spring stiffnesses available.

  The Stratus VSP consists of a viscoelastic compression ring seated within a telescoping housing. There are five stiffness rings available (very soft, soft, medium, firm, very firm), accommodating subjects from 90 to 250 lbs; we tested all five of the rings.

  The AirStance VSP is a pneumatic cylinder that suspends the patient's weight on pressurized air, accommodating persons weighing up to 300 lbs. We assumed three different body weights (BW) for adult subjects within the range of values recommended by the manufacturer and set the air pressure in the AirStance equal to 1/2 BW psi, and 1/2 BW±25%BW psi.

PROGRESS--Mechanical characterization of the three VSPs has been completed.

RESULTS--The three demonstrated substantially different static and dynamic properties. Static testing revealed that the stiffness (i.e., slope of force-displacement curve) for the Re-Flex was fairly linear, while for the AirStance and the Stratus the stiffness was nonlinear. All of the VSPs demonstrated hysteresis in the force-displacement plots, indicative of energy loss. Dynamic testing showed that the Re-Flex and the Stratus may be approximated with 2nd-order underdamped systems; the telescoping tube assembly of the Re-Flex may contribute to Coulomb damping. The dynamic response of the AirStance tended to be overdamped for lighter loads, but had overshoot and oscillations at the heavier loads; the total displacement of the AirStance during the step-load trials was relatively small, due to the extremely high stiffnesses in the range of loads tested.

FUTURE PLANS--We are about to begin a clinical investigation of persons with unilateral transtibial amputation walking with vertical shock pylons. Such subjects will be recruited for gait analyses in which they will walk with and without a VSP, while kinematic measurements and ground reaction forces are acquired. By comparing gait measurements for both conditions, we can determine whether or not the devices significantly affect the pattern of walking. The gait data will also be compared with that of controls to determine whether the shock-absorbing mechanisms enable subjects to walk more normally.




Vern L. Houston, PhD, CPO; Carl P. Mason, MSBE; Aaron C. Beattie, BS; Kenneth P. LaBlanc, BS, CPO; Gangming Luo, PhD; MaryAnne Garbarini, MA, PT; Cathy M. Cruise, MD
New York University Medical Center, New York, NY; VA Medical Center, New York, NY 10010; email: vlh3@is2.nyu.edu

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #A514-4DA)

PURPOSE--The objectives of this project are: to continue refinement and enhancement of the VA-Cyberware Prosthetics-Orthotics Optical Digitizer, and to conduct fundamental application studies with the optical digitizer to test and demonstrate its capabilities, effectiveness, and efficiency in quantitatively characterizing the spatial geometry and surface topography of residual limbs and limb and body segments.

METHODOLOGY--To achieve these objectives, the following research protocol was established:

  1. Develop design specifications for, and procure and test a second generation, prototype optical digitizer, correcting the deficits identified in the first, extending its capabilities, and improving its performance;
  2. Enhance and optimize the control, data acquisition, processing, visualization, and analysis software developed for the original prototype digitizer, integrating the modules into a single, user-friendly, menu-driven program for use by prosthetics-orthotics practitioners;
  3. Refine and optimize the control, tool path clearance, and the surface contour interpolation and smoothing software for the VA Prosthetics-Orthotics CAM milling machine to enable WYSIWYG prosthesis-orthosis design and manufacture from optical scan measurements;
  4. Continue development of lower limb prosthetics computer-aided socket design (CASD) templates for use with optically digitized residual limb measurements;
  5. Develop CAD templates for design of ankle-foot orthoses (AFOs) and knee-ankle-foot orthoses (KAFOs) from optically digitized limb segment measurements.

PROGRESS--Construction of a second generation optical digitizer prototype has been completed, after several unpredictable, protracted delays in production. Installation and calibration of the digitizer was begun, precisely measuring how its camera pixels map into the spatial field of view, from which algorithms will be derived to correct scan measurements for optical nonlinearities in the system. Rigorous laboratory and clinical testing of the digitizer will follow. Work continued on adapting and optimizing the digitizer control, measurement acquisition, processing, visualization, and analysis software, integrating it into a single, user-friendly, menu driven program running on a WindowsNTTM PC workstation for prosthetics-orthotics clinical practitioners. Enhancement and optimization of the control, tool-path clearance, surface contour interpolation and smoothing software for the VA Prosthetics-Orthotics CAM milling machine was completed. Development and refinement of CAD system prosthesis and orthosis design templates based on optically digitized residual limb/limb segment measurements also continued.

FUTURE PLANS--After testing and refinement, further application studies will be performed on the digitizer, including development of torso scanning procedures for CAD/CAM of thoracic-lumbosacral orthoses (TLSO); compilation of a quantitative prosthetics and orthotics patient database of residual limb/limb segment geometries, measurements, and histories for use in developing improved prosthetic socket and orthosis designs; compilation of a database of patient limb segment contours, areas, and volumes for correlation with, and quantitative assessment of, the efficacy of medical treatments and rehabilitation regimens; and utilization as an educational tool for direct visualization of prosthetics-orthotics principles and practices.




David A. Boone, CP, MPH; Douglas G. Smith, MD; James Christian Beck, MSME; David E. Mathews, CP
Prosthetics Research Study, Seattle, Washington 98122;

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #A2130-RA)

PURPOSE--The objective of this study is to determine the amount of flexibility subjects with transtibial amputation prefer in three planes and whether they consistently will prefer that amount over time.

METHODOLOGY--Twenty-five subjects of similar walking ability (Medicare class 3) will evaluate the compliance of an adjustable pylon/ankle while traversing a standardized course representing normal gait activities such as ascent and descent of stairs and a ramp. There will be three data collection sessions per subject spaced at 2-week intervals.

  A test prosthesis with a relatively rigid foot, an adjustably flexible ankle, and a rotator has been constructed. The prosthesis allows independent adjustment of the rotational stiffness in each plane and is instrumented to measure the displacement in each plane during walking. Force and moment will also be measured.

PROGRESS--A reliable, easily adjusted flexible ankle, quite similar to our original design concept, has been designed, constructed, and tested on three subjects. The instrumentation to measure force and displacement has been built and tested and works satisfactorily. Rather than LVDTs, we are using miniature cable transducers to measure displacement because of the difference in size and weight. A highly linear potentiometer is used to measure the axial rotation provided by an Otto Bock rotator. The test course has been constructed including the photosensor instrumentation to monitor velocity and position. Software for the collection of data and the subsequent analysis has been completed. Improvements that will make data collection more robust and free from error have been made to the telemetry. Subjects are currently being enrolled for data collection.




Alan R. Morris, MASc, PEng; Wes From, MASc, PEng; John Hancock; Devon Ostrom, MASc; Danial Cribbs, CP(C); Stephen Naumann, PhD, PEng
Bloorview MacMillan Centre, 350 Rumsey Road, Toronto, ON M4G 1R8; email: naumann@utcc.utoronto.ca

Sponsor: Industrial Research Assistance Program, National Research Council of Canada; Lever Foundation; Ontario Rehabilitation Technology Consortium funded by the Ontario Ministry of Health, Variety Ability Systems, Inc.

PURPOSE--The purpose of this research project is to develop a running prosthesis that will enable smooth and efficient gait for children between the ages of 5 to 12 with an above-knee amputation.

PROGRESS--Work has continued since the creation and testing of the original prototype knee. A second design has been created with a separate component that provides relief of shock during fast weight acceptance. A prototype of this shock absorber has been created for client evaluation. The new running knee prosthesis is currently undergoing component evaluation for strength and cycle life, and a number of new features have been integrated into it to maintain market relevance.

FUTURE PLANS--Upon completion of the component evaluation for the new design, a working prototype will be constructed and this second prototype will be tested on clients at the Centre.



Micaela Schmid, DrEng; Daniela Zambarbieri, DrEng; Gennaro Verni, DrEng
Dipartimento di Informatica e Sistemistica, Università di Pavia, Via Ferrata 1, 27100 Pavia, Italia; Centro protesi INAIL, Via Rabuina 14, Vigorso di Budrio (BO), Italia; email: miki@linus2.unipv.it; dani@unipv.it; cprotesi.budrio@inail.it

Sponsor: Centro Protesi INAIL via Rabuina 14, Vigorso di Budrio, Bologna, Italy; Istituto Superiore di Sanità, Viale Regina Elena 299,00161 Roma, Italy; email: cprotesi.budrio@inail.it

PURPOSE--An internal model of the human body is used by the central nervous system to decide the adequate motor commands needed to execute movements. The presence of lesions, such as limb amputation, induces a mismatch between the output predicted by the internal model and the movement actually executed by the body. Thus, a reorganization of the motor strategies is needed that induces an update of the internal model. In prosthetic subjects, rehabilitation can induce the update of the internal model. If the subject is provided with some kind of artificial sensory reafference, it is likely to assume that the process of updating the internal model can be improved. The aim of our research is to develop a system especially designed to provide sensory biofeedback to subjects with lower limb amputations.

METHODOLOGY--This system provides the subject with information about the position of his center of pressure (CP) during the gait. The system is made of two major components: sensorized insoles to detect foot pressure on the ground, and a tactile stimulation device. Each insole is made of a matrix of 64 pressure-sensitive sensors. The output signal, provided by each sensor, is locally processed by means of a small resistive circuit connected to the insole. The circuit provides in real time the X and Y coordinates of CP position and the vertical component of the resulting force. The tactile stimulation is realized by four small eccentric motors placed on the thigh or trunk, depending on the level of amputation. The aim of this biofeedback is to inform the subject when his CP falls outside a specific area that represents the right load on the foot, during all phases of gait. This area will be referred to as the "normality area."

PROGRESS--In order to define a method for the evaluation of the normality area we have examined ten control subjects without any postural or locomotion problem. The subjects walked for 20 steps wearing a pair of sensorized insoles under three different conditions: bare foot, with leather shoes, and with gym shoes. The spatial value that CP assumes during walking changes with each step, resulting in a set of trajectories that are very similar and close to each other, covering a specific surface of the foot sole corresponding to the normality area previously described. The outline of this area is evaluated by computing first the mean trajectories ±2.5 SD and then by a best fitting with a fourth degree polynomial function. Moreover, we have observed that the normality area is strictly related, at least in controls, to the type of shoes worn.

FUTURE PLANS--The first step will be to calculate the normality area for different classes of lower limb prosthetic subjects, depending on the level of amputation and the prosthesic components. The second step will be that of applying this system during the rehabilitation program in a selected population of lower limb amputees at the Centro Protesi Inail, Budrio, Italy.




Vern L. Houston, PhD, CPO; Carl P. Mason, MSBE; Luigi Arena, MD, PhD; Gangming Luo, PhD; Kenneth P. LaBlanc, BS, CPO; MaryAnne Garbarini, MA, PT; Cathy M. Cruise, MD
New York University Medical Center, New York, NY; VA Medical Center, New York, NY 10010; email: vlh3@is2.nyu.edu

Sponsor: G.T.H. LAMB Group, New York, NY

PURPOSE--The purpose of this project is to investigate the biomechanical and vascular characteristics and geometry of residual limb tissue and quantify prosthetics loading, so fundamental socket design principles and algorithms can be derived that afford more intimately fitting, comfortable, and functional sockets for persons with lower limb amputation.

METHODOLOGY--The following protocol for a representative sample of physiologically mature and prosthetically habituated persons with transtibial amputation is being conducted:

  1. Measure and digitize residual limb spatial geometry of the subjects, its tissue biomechanical characteristics, and principal vascular geometry;
  2. Measure socket/limb interface stresses during stance and gait;
  3. Measure residual limb principal vascular flow velocities and volumes, without and with sockets on under axial loads;
  4. Derive 3-D, nonlinear, viscoelastic finite element (FE) models of the residual limbs;
  5. Validate the FE models by comparing respective analysis-predicted tissue surface stresses and subsurface displacements at key anatomical locations with corresponding empirically measured socket interface stresses and magnetic resonance (MR) measured tissue displacements, refining the FE models as necessary;
  6. Calculate the effects of various prosthetic socket design geometries and prosthetics loads on socket/limb interface stresses, tissue strains, and vascular perfusion velocities and volumes using the (refined) FE models developed for the subjects;
  7. Analyze the information and measurements compiled, and formulate socket design principles therefrom, as a function of patient tissue biomechanical characteristics, vascular state, and applied prosthetic loads;
  8. Develop CAD socket design algorithms implementing the formulated design principles, and clinically test and measure the degree of fit, comfort, and function provided by the resulting CAD designs.

PROGRESS--An electromechanical, servo-actuated indentor with force/position feedback was constructed, and the biomechanical creep and stress relaxation responses of the limb tissues of 32 test subjects measured, and subsequently mathematically modeled, using large deformation strain energy material models, with viscoelastic exponential transient and Ogden elastomeric steady-state components. The residual limbs of three of these subjects, representative of their respective types: one with soft, one with average, and one with firm durometer tissues, were digitized with high-resolution optical and MR scanners. The data compiled were used to create FE models of the residual limbs consisting of membrane elements of the integument, hexahedral elements of the underlying soft tissues, and rigid body elements to define bones. The subjects' well-fitting, habituated sockets were electromechanically digitized and modeled using stiff elastic hexahedral elements with moduli derived from material mechanical tensile tests. Static and dynamic socket/limb interface stresses were measured with VA-Tekscan P-Scan transducers for the subjects in their original, well-fitting habituated sockets, and in two experimental sockets with popliteal depression and posterior brim geometric design variations, but otherwise identical to the subjects' original sockets. MR high-resolution morphological and angiographic scans of the limbs, and flow velocity and volume scans of their popliteal arteries in each of the respective sockets were acquired with 32 Kg static, axial loads applied. Corresponding FE simulations of application of each socket, and subsequent axial loading were performed. The resulting FE analysis predicted residual limb surface stresses, and subsurface tissue displacements at key anatomical locations were calculated and compared with P-Scan measured socket/limb interface stresses and MR measured tissue displacements. Reasonably close agreement was obtained between the predicted results and empirically measured tissue stresses and displacements, demonstrating the utility and effectiveness of FE analysis as a prosthetics design and analysis tool. Investigation of the effects of variations in other socket design geometries on residual limb tissue stress and strain and vascular perfusion is continuing.




Deanne Gusdal, BSc; Stephen Naumann, PhD, PEng; W. Cleghorn, PhD, PEng
University of Toronto; Bloorview MacMillan Centre, 350 Rumsey Road, Toronto, ON M4G 1R8, Canada; email: naumann@utcc.utoronto.ca

Sponsor: Natural Sciences and Engineering Research Council

PURPOSE--The currently available prosthetic alignment device for persons with lower limb amputation is of a size and weight that is too large for use with children. A small, lightweight force transducer, suitable for use with children 7 to 12 years of age, is therefore being designed to aid the prosthetist in the alignment process.

  This transducer should increase the efficiency of the procedure, reduce the amount of experience needed by the prosthetist to achieve an acceptable alignment, and thus reduce the overall cost of the process. It will also result in an optimum alignment in a shorter period of time, thus increasing the satisfaction of the client.

PROGRESS--To date, discussions with a prosthetist have been held to determine the age group and size constraints within which this device must fit. The forces that the device must be capable of withstanding when in use are being determined to aid in the selection of an appropriate, lightweight material for construction of the transducer. A finite element analysis is being performed to assure the device will withstand the forces applied by children. This analysis will also be used to determine the ideal shape and configuration necessary to achieve the desired measurements from the transducer.

FUTURE PLANS--Once the optimum material and configuration for the force transducer have been determined, a prototype will be manufactured to test the design. This prototype will be used to ensure the device can withstand the loads it has been designed for, as well as to determine the calibration matrix for the transducer. The calibration matrix will be needed to convert the measurements the device reads into the information the prosthetist requires to perform the alignment adjustments. The prototype will then be tested during the alignment of a child's prosthesis to ensure it is suggesting the changes needed to achieve the correct alignment.



Stephen Naumann, PhD, PEng; Alan R. Morris, MASc, PEng; Kimberley Parker, MASc, PEng; Wes From, MASc, PEng; Dan Cribbs, CP(c); Mary Huggins, BSc, PT
Bloorview MacMillan Centre, 350 Rumsey Road, Toronto, ON M4G 1R8, Canada; email: naumann@utcc.utoronto.ca

Sponsor: Ontario Rehabilitation Technology Consortium (funded by the Ontario Ministry of Health)

PURPOSE--Alignment of a lower limb prosthesis involves positioning its socket relative to other components such as knee joint and foot. The prosthetist observes how the client walks and inquires about the level of comfort to determine an appropriate alignment. This process involves a fair amount of refined repetition to provide an optimum alignment and is a skill that the prosthetist gains over time. The objective of the Dynamic Prosthetic Alignment Assistant is to provide guidance to clinicians on alignment of lower limb prosthetic components based on quantitative information obtained during a clinical fitting. A load cell placed between the socket of the client and prosthetic components will obtain the magnitude and direction of the weight-bearing forces. A computer program will record and visually display the information while the client walks, and recommends appropriate changes to obtain an optimal alignment.

PROGRESS--A prototype alignment assistant and data collection software has been developed. The prototype contains a load cell (developed by an outside company) capable of interfacing with prosthetic limb components. To fully develop the analytical relationships between the alignment changes and quantitative measures, a clinical study into the effects of alignment changes on gait was initiated. In this study, kinematic information was obtained, along with the prototype weight-bearing forces, during walking for various alignment configurations. The study focused on three persons with transtibial amputation. Initial data for good alignment was obtained from the prosthetist. Alignment changes between the socket and the pylon were made looking at changes in linear displacements and tilts between the components from the good alignment. This has allowed for an investigation on the sensitivity of quantitative measures to alignment changes as well as investigation of the effectiveness of applying simple biomechanical relationships. Few to no differences were found in the kinematic variables between alignments, indicating that the range of motion of the prosthetic limb is not significantly affected by alignment changes. Significant differences were found between measured moments for various alignment changes.

FUTURE PLANS--Data collection for various alignment configurations will continue, and the number of subjects in this study will continue to expand. Analysis of these data will enable us to develop the analytical relationships between the alignment changes and the measured parameters. Work will continue on the development and refinement of the software prototype for use by the clinicians. This will involve testing of various user interface setups with in-house clinicians.



David A. Boone, CP, MPH; Douglas G. Smith, MD; Kim L. Coleman, MS; Linda S. Laing, BA; David E. Mathews, CP
Prosthetics Research Study, Seattle, Washington 98122; email: David@prs-research.org; Doug@prs-research.org; Kim@prs-research.org; Linda@prs-research.org;

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #A2122-RA)

PURPOSE--This study had two primary objectives: to quantify differences in daily ambulatory activity when persons with transtibial amputation wore a "low-end" versus a "high-end" prosthetic foot, and to quantify differences in subjects' response to, and satisfaction with, the transtibial prostheses when wearing the two different prosthetic feet.

PROGRESS--This project was completed March 1998.

RESULTS--Thirty-eight subjects (30 males, 8 females) with transtibial amputation, who functioned at or above Medicare DMERC activity level classification 3, and who had no major medical or musculoskeletal complications were enrolled. Potential subjects anticipating a lifestyle altering event--such as change of job, marital status, residence, or number of household members--were excluded. All subjects were at least 1-yr post-amputation. Thirty-two (29 males, 3 females) completed the study. Two ceased participation with no explanation, and one each did so because of auto accident, congestive heart failure, dislike of study foot, or change of residence.

Measurement Tools
  The Step Activity Monitor (SAM) is a small microprocessor-controlled instrument that is worn on the ankle. It continuously records step activity at adjustable intervals (e.g., 1 min, 2 min, etc.). SAM measures recorded in this study included: average steps/day; % time spent at low activity (1-15 steps/min); % time spent at moderate activity (16-30 steps/min); % time spent at high activity (>30 steps/minute); and hours of inactivity.

  The Prosthetics Evaluation Questionnaire (PEQ) is a validated instrument used to assess patients' experience with their prosthesis. The PEQ quantifies differences in patient satisfaction, and identifies the source of the differences in terms of prosthesis-specific issues and characteristics. The questions are combined into the nine scales, the scores of which are used as the outcomes measures. The scales are: ambulation, residual limb health, utility, appearance, sounds, frustration, perceived response, social burden and well being.

  Each subject completed two randomly ordered 5-week trials, one with each study foot. During each trial, the subject was fit with a foot (1 week), accommodated for 2 more weeks, then was monitored with the SAM (recording step counts every minute) for 2 weeks. Alignment of the prosthesis was set optimally for each foot rather than maintained constant throughout all conditions. At the beginning and end of each monitoring period, measures of average step length were collected, and an outdoor walking trial was performed to measure the accuracy of the SAM for step counting. The PEQ was administered at the end of each SAM-monitoring period. After the last monitoring period and while still blinded, subjects were asked their preference between the two study feet. At this point, the subjects were also asked which foot (original, SACH, FlexWalk II) they wanted for continued wear. The identity of the feet was revealed after selection. Subjects were allowed to change their choice of foot for continued wear if desired.

Methodological Validation
  A test for carryover effect was performed to determine whether the order of conditions affected step activity. Two-tailed heteroscedastic analysis indicated that the order of conditions did not influence results (p=0.1748). This result indicates that the cross-over study design was effective and the data can be treated accordingly.

Step Activity Results
  Step activity results showed significant differences between the study feet in step length, minutes spent at moderate level activity, and estimated distance walked per day. Because the difference in step length between the two feet that showed a high degree of statistical significance, the step count data were scaled by step length to achieve an estimated distance walked per day. It is interesting to note that the unscaled measure of overall activity (steps/day) did not show a difference at the p<0.05 level, while the scaled data did show a significant difference. This 0.55 km difference corresponded to a 9.5 percent increase in estimated distance walked with the FlexWalk II foot, relative to the SACH foot. For all step activity measures showing differences, the FlexWalk II was associated with increases in activity.

  The accuracy of each step monitor for each subject was evaluated before and after every 2-week monitoring trial. Accuracy was determined by comparing the average of two observers' counts for a two-block walking trial to the count recorded in the SAM memory. The absolute value of the percent error for each trial of each subject was used to calculate the accuracy statistics. Overall accuracy across the study was 99.41 percent.

PEQ Results
  Two scales of the PEQ showed significant differences between the study feet: the Perceived Response Scale and the Utility Scale. These differences were in the same direction as the step activity results with the FlexWalk II foot being associated with improved scores. (p<0.05)

Preference between All Feet
  When asked to compare the two study feet, 22 subjects preferred the FlexWalk II foot, 9 preferred the SACH, and 1 had no preference. When asked to select their preferred foot for continued wear, 14 subjects chose the FlexWalk II, 4 chose the SACH, and 14 chose their original foot.

  All subjects were still blinded as to the identity of the study feet when making their selection. The strength of the preference varied considerably between subjects; some perceived differences in great detail and emphatically expressed preference, while others stated that differences seemed small and not of great importance

IMPLICATIONS/FUTURE PLANS--This trial has demonstrated that statistically significant differences can be measured between the Flex Walk II and the SACH foot. These differences reflect greater physical function and subjective satisfaction with the Flex Walk II overall. However, a considerable number of subjects stated their preference for the SACH foot indicating the continued value of this design for many persons with amputation.

  A similar protocol utilizing the SAM and the PEQ is being used to investigate the differences between PeLite socket liners with neoprene sleeve suspension and an elastomeric liner with distal licking suspension pin.



E. Lower Limb: Transfemoral



Tracy Orr, PhD; Myron Spector, PhD
Brockton/West Roxbury VA Medical Center, West Roxbury MA 02132; Brigham & Women's Hospital, Boston MA 02115; email: spector@ortho.bwh.harvard.edu

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #A498-DA)

PURPOSE--The objective of this investigation is to determine the optimal surface characteristics and materials properties for a femoral endoprosthesis to avoid loosening. We seek to finalize the femoral stem design based on the results from the finite element analyses and mechanical testing of our earlier work, and have the implant manufactured to our specifications. There are two stages to the development of our femoral component. The first will validate our finding that the optimum porous coating distribution for a canal-filling titanium component is one that goes beyond the metaphyseal region of the femur, into the diaphysis (approximately 60 percent of the available surface area). We will begin preparation to start limited clinical trials using this prosthesis. The second stage is the development of the metal core-polymer shell composite. We will finalize this design using our optimization algorithms.

  We also seek to perform mechanical testing that will be used as groundwork for human implantation of the implant. The porous-coated titanium alloy stem will be tested under static loads and cyclic loading to determine the stiffness, ultimate strength, and fatigue strength (endurance limit) of the device. Cadaveric studies will be performed to determine the amount of micromotion and stress shielding using the procedure developed in the prior grant period.

METHODOLOGY--This project investigates PMMA and PEEK as possible shell materials. The shear strength of the PMMA/stainless steel interface is experimentally determined through torsion, pushout, and fatigue testing for both rough and smooth metal cores. All testing was carried out in an environmental chamber containing water at human body temperature. This chamber was designed and manufactured as part of this research.

PROGRESS--Specimens of PMMA- and PEEK-coated metallic rods have been fabricated and are undergoing mechanical testing to determine the inter-facial strength of attachment.

RESULTS--For relatively young and active patients, total hip replacement has a poor long-term prognosis. For them, loosening of the implant is the most common complication, generally due to a failure of the interface between the bone and the prosthesis, or between the bone and the bone cement for cemented components. The problem to be addressed in this study is the failure of femoral components due to the loss of fixation and stress adaptive bone remodeling.

  The method of this study provided all the tools necessary to solve the optimization problem. Qualitatively, the trends in the data suggest that the optimal distribution (which minimizes the combination of stress shielding and relative motion) is to coat somewhere where the proximal 60 and 81 percent of the stem length. These conclusions supported many design principles for cementless prostheses that are based on clinical, analytical, and experimental studies. The experimental study was unique, however, in its development of an experimental model for studying the variation in surface coating distribution and in its experimental demonstration of the complicated interaction between relative motion and surface strains with variation in bonding distribution. The conflicting effects of varying the surface coating distribution on the two parameters confirm that an optimization approach is necessary to determine the most suitable coating distribution, where the optimal solution corresponds to a compromise between the parameters.



Tracy Orr, PhD; Myron Spector, PhD
Brockton/West Roxbury VA Medical Center, West Roxbury MA 02132; Brigham & Women's Hospital, Boston MA 02115; email: spector@ortho.bwh.harvard.edu

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #A2116-2RA)

PURPOSE--The primary purpose of this study is to develop a classification algorithm to assess fracture risk in femurs with total hip replacement prostheses when osteolytic lesions due to wear particles are present. Both size and location of the lesions are considered. A further purpose is to evaluate current treatment methods for these wear-particle-induced osteolitic lesions.

METHODOLOGY--A two-pronged approach is undertaken in setting up a model for the osteolytic lesions. In the first approach, broadly labeled the experimental approach, the following method is adopted. Clinical radiographs are used to determine the size, geometric description, and location of the osteolytic lesions. Lesions will be modeled in cadaver femurs, by introducing defects into the specimens. The femurs with the introduced defects will be radiographed and compared to the clinical radiographs to ensure reasonable starting point for the modeling. Typical procedures will be used to implant the hip prostheses into the femurs containing the defects. Next, the cadaver bone models will be mechanically tested in a loading rig. Typical values such as the loading force, cross head displacements, and surface strains will be measured.

  In the second approach, a numerical study of the fracture risk of femurs with the osteolytic lesions will be undertaken. The finite element method will be used to model the femurs with the hip implants. Typical quantities such as applied force versus global displacement, surface strains, and stresses, especially around the lesions, will be calculated. Wherever possible, the numerically predicted results will be compared with the experimentally measured values. This will verify the numerical results.

  Once the finite element model is verified, it can be used for predictive purposes. In particular, the magnitude of stress intensification will be calculated due to the introduced lesions/defects.

  Parametric studies, in which the size and location of the osteolytic lesions are varied, will be performed. The fracture prediction will be made for each case of the parametric study. The classification of fracture susceptibility will be based on this parametric study.

PROGRESS--On the experimental front, we have reviewed clinical radiographs and have classified the lesions based on their locations, size, and the length of the implanted femoral stem. Further, a number of defects were generated experimentally in cadaveric femurs. The results were radiographed and compared to the clinical radiographs. The comparison was done after all radiographic images were digitized. Also, the loading rig housing was built and preliminary tests were conducted. We stand ready to commence with the first mechanical tests of the control case.

  On the numerical modeling front, we have completed a limited and preliminary parametric study. Two loading cases were considered, with loads corresponding to stair climbing and single leg stance. A simplified and idealized geometry model was considered in this study. Lesions of various sizes and at various locations were modeled. A preliminary stress measure was defined and computed for all simulations performed.

RESULTS--The preliminary numerical model showed that, as expected, the presence of the lesions increase the stresses in the bone. Also as expected, the stresses increased as the size of the lesions increased. Lesions which penetrate the cortex 80 percent can increase the stresses by approximately 70 percent when compared to the control case. A caveat is in order: these results are preliminary and require further verification.

FUTURE PLANS/IMPLICATIONS--The preliminary numerical model will be revisited. A more accurate geometric representation of the experimental bone(s) will be made. More economic elements will be used than in the preliminary study, thus allowing for greater refinement to be made closer to the lesions, but still to maintain a reasonable model size. Further and more advanced stress measures will be developed in order to compare the cases with lesions against the control case(s).

  The information learnt from the preliminary parametric study will be used to design and implement more sophisticated studies in the future.

  On the Experimental front, the control case will be mechanically tested. The results will be compared with the Numerical model. Realistic lesions will be introduced and the femurs will be mechanically tested.



Gary M. Stefanov, BASc, PEng; Stephen Naumann, PhD, PEng
Institute of Biomedical Engineering, University of Toronto; Rehabilitation Engineering Department, Bloorview MacMillan Centre, 350 Rumsey Road, Toronto, ON M4G 1R8, Canada; email: naumann@utcc.utoronto.ca

Sponsor: Natural Sciences and Engineering Research Council

PURPOSE--It is important for the social development of children with above-knee amputations that they not be sheltered from their environment, and, as much as possible, take part in the same activities as their peers. These activities include playing in water while on the beach, in the pool, or at the playground. Most conventional prostheses will not stand up to repeated usage in water and related elements, such as sand, chlorine, and salt. The focus of this project is to design and develop a simpler, more easily manufactured knee joint for use in a pediatric swimming prosthesis. The device will have the capability to lock during gait, but can be unlocked to provide knee flexion capability to allow the child to sit or play at floor level.

PROGRESS--Work to date has focused on outlining the design requirements for the knee joint and determining the conceptual configuration of the eventual component. Design loads have been investigated, based on anthropometric data for the heights and weights of the children in the target population, as well as their expected level of activity. Detailed investigations of the environments in which the knee joint will be required to operate have been completed to identify the conditions which the materials and construction must withstand without functional degradation. Immersion tests of polymers in a chlorinated water environment and material strength tests have been undertaken to evaluate various chemically resistive polymer materials for use in the swimming knee application.

FUTURE PLANS--Currently a 3-D solid model of the proposed design is under construction. This modeling will allow the design to be optimized on previously identified design criteria. Optimization of the proposed design will aid in determining the final dimensions of the various structural components incorporated in the current configuration. With final dimensions of the knee joint components, manufacturing drawings will be produced in preparation for manufacture of a prototype knee joint unit. Subsequently, a prototype knee joint will be manufactured for bench testing, as well as for functional testing with an appropriate subject. Outcomes of bench tests, functional tests, and the generated user feedback will identify appropriate design modifications required to maximize the utility of the swimming knee. The anticipated time for completion of the project is fall 1998.


F. Lower Limb: Transtibial



Vern L. Houston, PhD, CPO; Carl P. Mason, MSBE; Kenneth P. LaBlanc, BS, CPO; Aaron C. Beattie, BS; MaryAnne Garbarini, MA, PT; Gangming Luo, PhD
New York University Medical Center, New York, NY; VA Medical Center, New York, NY 10010; email: vlh3@is2.nyu.edu

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #2068-RA))

PURPOSE--The objectives of this project are to update the design of, and the software for, the VA-Cyberware BK Prosthetics Optical Digitizer, and to establish the performance, repeatability, consistency, and temporal efficiency achievable with the digitizer in clinical prosthetics measurement of the residual limb spatial geometry and surface topography in persons with transtibial amputation.

METHODOLOGY--To achieve these objectives, the following research protocol has been established:

  1. Update the design of, and the control, measurement acquisition, and processing software for, the VA-Cyberware BK Prosthetics Optical Digitizer;
  2. Procure, calibrate, and laboratory test five enhanced design digitizer prototypes;
  3. Deploy the digitizer prototypes at five VA Medical Center Prosthetics CAD/CAM Laboratories, training the respective clinical prosthetics staffs in their operation and use;
  4. Conduct a clinical trial of the optical digitizer at the five VA Medical Center Prosthetics CAD/CAM laboratories, to establish the level of training and technical expertise required to effectively operate the digitizer; the degree of repeatability, consistency, and temporal efficiency afforded by the optical digitizer in measuring BK amputee residual limb spatial geometry and surface topography, especially in comparison with current CAD plaster wrap cast/electromechanical digitization techniques; and the level of durability, maintainability, and reliability of the digitizer in clinical settings;
  5. Compile, analyze, and document the results of the clinical tests, to identify any further design refinements and/or enhancements required before the digitizer is released for commercial production and general clinical deployment.

PROGRESS--Specifications for upgrade of the optical digitizer were developed, and the hardware updated with more advanced, state-of-the-art, optoelectromechanical components and mechanical design enhancements to make the digitizer more durable with lower maintenance requirements. An initial enhanced prototype was procured, and requisitions for four additional units for clinical testing at four other VA Medical Center Prosthetics CAD/CAM laboratories were placed. We have developed instrumentation for automated calibration and laboratory testing of the digitizers prior to clinical deployment. Work continued on adaptation and optimization of control, data acquisition, measurement processing, and analysis software, as well as that for automated landmark detection, identification, and registration, for use with the new digitizers. Preparation of instructional and operational manuals for the digitizer also continued.

  Laboratory testing and calibration will be conducted upon receipt of the prototype digitizers. The units will then be shipped and installed, and the prosthetics staffs trained, at the four prosthetics CAD/CAM laboratories. Clinical field testing of the digitizers with five subjects at each site shall then be performed. The test data compiled in the digitizer field trials shall be analyzed and recommendations made for incorporation of any enhancements and design refinements found to be appropriate.

FUTURE PLANS--Further application studies with the digitizer shall be pursued, including the compilation of a quantitative database of residual limb geometries, measurements, and histories for use in developing improved prosthetic socket designs; the compilation of a database of residual limb contours, areas, and volumes for correlation with, and quantitative assessment of, the efficacy and temporal efficiency of post-surgical treatment and rehabilitation regimens; and the utilization of the digitizer as an educational tool for enhanced visualization and analysis of prosthetics principles and clinical practices.




Vern L. Houston, PhD, CPO; Carl P. Mason, MSBE; Kenneth P. LaBlanc, BS, CPO; Aaron C. Beattie, BS; Gangming Luo, PhD; MaryAnne Garbarini, MA, PT
New York University Medical Center, New York, NY; VA Medical Center, New York, NY 10010; email: vlh3@is2.nyu.edu

Sponsor: Department of Veterans Affairs, VA Rehabilitation Research and Development Service, Washington, DC 20420
(Project #2220-RA)

PURPOSE--Our objective is to evaluate the capabilities of, and measure the clinical performance achievable with, the hand-held digitization and virtual casting systems that have been recently developed for the characterization of residual limb and limb segment spatial geometry and surface topography.

METHODOLOGY--To achieve this objective, we have established the following research protocol:

  1. Procure commercially available models of hand-held digitizers, and obtain the instruction and training in their operation and use provided by the manufacturers;
  2. Conduct laboratory tests of the digitizers to determine the maximum achievable accuracy and repeatability, the susceptibility to extraneous electromagnetic and optical interference, and the maintenance requirements, durability, and reliability of each system;
  3. Conduct clinical tests with the digitizers to determine: the level of knowledge, training, and skill required to effectively operate them, the maximum and the average accuracy clinically achievable (as well as the average intra-prosthetist repeatability and inter-prosthetist consistency achievable with each digitizer in characterizing patients' residual limb/limb segment spatial geometry and surface topography, including detecting and identifying predesignated limb anatomical landmarks) with them; and the average time required to scan a residual limb, input the measurement data into a CAD system, and design a socket to specified dimensions and shapes with them;
  4. Analyze, document, and disseminate the project results.

PROGRESS--Commercially available hand held prosthetics-orthotics digitizers were obtained from BioSculptor Corp., CAPOD Systems, Inc., Seattle Limb Systems, and Tracer Corp. for evaluation. Training and instruction in operation and use of the systems from each of the respective system manufacturers was begun, as was laboratory testing and evaluation of the digitizers' accuracy, repeatability, maintenance requirements, and susceptibility to electromagnetic interference. Further laboratory testing of the system's performance, susceptibility to interference, and durability and reliability will continue. Rigorous clinical tests of the systems to determine the average performance, effectiveness, and efficiency of each system in a clinical environment will be conducted in the near future.



Joseph Czerniecki, MD; Glenn Klute, MS; Blake Hannaford, PhD
VA Puget Sound Health Care System, 1660 S. Columbian Way, Box 358280, MS151, Seattle, WA 98108; Departments of Bioengineering and Electrical Engineering, University of Washington, Seattle, WA 98195-2500; email: czerniecki.joseph_m@seattle.va.gov; gklute@u.washington.edu; blake2@isdl.ee.washington.edu

Sponsor: VA Rehabilitation Research and Development Center for Excellence in Amputation, Prosthetics, and Limb Loss Prevention, Grant #A0806-C

PURPOSE--Kinematic and kinetic analysis of walking speed gait reveals the intact ankle supplies a significant amount of energy during stance phase plantar-flexion. This energy is unavailable to the amputee. The objective of the Powered Prosthetic Project is to improve the gait of persons with transtibial amputation by using a powered prosthetic limb to supply the missing energy.

METHODOLOGY--Our approach is to design and fabricate a powered, lower limb prosthetic device. We intend to use this device to supply energy during the gait cycle, and plan to measure and compare the performance of the powered prosthesis with conventional, passive devices. To measure our success in this endeavor, we have formulated three hypotheses regarding the powered prosthetic limb: it 1) will reduce the metabolic cost of locomotion, 2) will improve the symmetry of gait, and 3) reduce the perceived level of effort required for locomotion. Our first development milestone is the identification of prosthetic limb performance requirements. Once completed, we intend to develop an actuator that functions like the missing muscles (gastrocnemius and soleus) and tendon (Achilles) of the amputated limb. Following the design of an artificial musculotendon-like actuator, we will specify and tune a controller using numerical methods supplemented with bench-top hardware tests including actual applied loads and motions. After a committee review of performance tests, safety interlocks and fail-safe measures, we plan to conduct tests with human subjects.

PROGRESS--We are in the first year of development and have identified the performance requirements expected of a powered prosthetic limb. Following a trade study of potential actuators and power sources, we conducted laboratory tests of a custom-built, McKibben pneumatic actuator. The McKibben actuator is well suited to this application because of its high power-to-weight ratio. In an effort to incorporate as many muscle-like features as possible into the hardware, we have also designed a miniature hydraulic system intended to operate in parallel with the actuator. Numerical simulations of this system provide confidence that desirable Hill-like muscle damping can be achieved. To accommodate energy storage and release during the gait cycle, we surveyed the biomechanics literature to identify requirements of an artificial Achilles tendon. In a trade-off between performance and a penalty for excess weight, we have specified a two-spring implementation that we predict will achieve a performance comparable to a typical Achilles tendon.

FUTURE PLANS--We are now fabricating a bench-top test unit for laboratory performance tests. Our future steps include conducting artificial musculo-tendon performance tests, designing the control system, conducting hardware and software tests to insure robust performance, followed by human subject tests.



Bruce Sangeorzan, MD; Joseph Czerniecki, MD; Kelly Weaver, MD; A. McCormack; Randall Ching, PhD; David A. Boone, CP; Allan Tencer, PhD
VA Puget Sound Health Care System, 1660 S. Columbian Way, Box 358280, MS151, Seattle, WA 98108; Department of Orthopaedics, Biomechanics Lab, University of Washington, Seattle, WA 98195

Sponsor: VA Rehabilitation Research and Development Center for Excellence in Amputation, Prosthetics, and Limb Loss Prevention, Grant #A0806-C

OBJECTIVES--We seek to determine whether the constraint provided by a rigid orthosis alters the kinematics and position of the hindfoot, and the resultant pressure distribution to the forefoot.

METHODOLOGY--An experimental flatfoot model is used to determine the effects of rigid and compliant orthoses on the kinematics of the hindfoot, and distribution of forefoot pressures in twenty specimens.

  CT scans are obtained to delineate the bony architecture of each foot. Data are downloaded into software that creates 3-D images of each bone. Custom orthoses (rigid and compliant) are designed for each foot using a CAD/CAM system. Each specimen is placed on a test frame that simulates physiologic loading in heel-strike and stance phases of the gait cycle. Data are obtained on hindfoot kinematics using a 3-D motion tracking system. Plantar pressures through the foot are obtained simultaneously with a dynamic pressure measuring system placed between the foot and orthosis. Each specimen is tested in the intact and flatfoot condition, with and without each orthosis, in heel-strike and stance positions. Statistical comparison of hindfoot bone angular motion in three planes and pressure distributions throughout the plantar surface of the foot are determined.

PROGRESS--Twenty specimens have been prepared for testing. Preliminary radiographic and arch volumes have been obtained. CT scanning of the specimens has begun for the 3-D modeling. Casting for orthoses is complete and fabrications in process. Loading apparatus is being configured for computer control to allow greater precision and accuracy of measurements.

IMPLICATIONS--Foot ulceration secondary to conditions such as diabetes poses significant problems to individuals and a vexing challenge to treating practitioners. Gaining an understanding of potential causes of foot ulcers including increased pressures across the forefoot, bony malalignment, and changes in relative motions between bones can lead to a more systematic approach to treatment and prevention of this problem.



Justus F. Lehmann, MD; Robert Price, MSME; John Fergason, BS; Ramona Okumura, BS; Grace Koon, BS
University of Washington, Department of Rehabilitation Medicine, Seattle, WA 98195-6490; email: pricer@u.washington.edu

Sponsor: National Institute of Disability and Rehabilitation Research (NIDRR), U.S. Department of Education, Washington, DC 22202

PURPOSE--In mechanical systems, such as prostheses, the natural timing of energy storage and release is dependent on the mass, elastic stiffness, and damping of the system. At its resonant frequency, the system will store and release energy most effectively. This study tests the hypothesis that metabolic efficiency of ambulation is minimized in persons with transtibial amputation (TA) wearing a prosthetic "energy restoring" foot/ankle system when the driving frequency (dictated by ambulation speed) corresponds to the system's resonant frequency (dictated by foot/ankle stiffness and user mass).

METHODOLOGY--Three tasks are involved in this process: 1) selecting the appropriate foot/ankle stiffness to achieve desired resonance frequencies for a given user's mass; 2) measuring the driving frequency of the system during ambulation for each ambulation speed; and 3) determining the effect of the resonant to driving frequency ratio on energy efficiency. The Air-Flex prosthetic foot/ankle system permits stiffness adjustment through inflation of an air bladder engaging an auxiliary leaf spring that acts to increase the stiffness of the basic foot over and above the stiffness of the main structure. Four stiffnesses are tested on each TA subject. This is achieved by employing two basic main structures, each of which are tested at a bladder pressure of zero (uninflated) and 45 psi.

  The resonant frequency is determined on a test device to simulate in situ use by using weights to simulate subject mass. In addition, an impedance measurement technique is being explored that measures resonant frequency of the foot/ankle system while the subject wears the prosthesis. This will be accomplished by instrumenting the prosthesis with an accelerometer and tapping the foot with an impedance hammer while the subject stands on the metatarsal head area of the prosthetic limb. The input versus output response will provide the information necessary to determine the transfer function for the system and its resonant frequency under in-situ conditions. Measuring the driving frequency during gait is accomplished via strain measurements, with strain gages mounted on the foot/ankle system of the prosthesis. The driving frequency for each stiffness, at four ambulation speeds (80, 100, 140, and 180 m/min), is measured during treadmill ambulation.

  During the ambulation, metabolic testing is also performed, using a computerized test cart. The metabolic testing protocol consists of a 2-min rest period, from which a metabolic baseline is determined, a 2- to 4-min exercise period, and a recovery period. For the two walking speeds (80 and 100 m/min), a 4-min exercise period is used; for the two running speeds (140 and 180 m/min.), a 2-min exercise period is used. Oxygen consumption is measured and averaged over 30 s intervals. Metabolic cost is determined by the oxygen consumed over and above the baseline, throughout the test, including the recovery period. The net quantity of oxygen consumed is normalized to the subject weight and the distance ambulated to arrive at a metabolic efficiency value (ml O2/kg/m). The relationship of the ratio of resonant to driving frequency as it relates to metabolic efficiency is determined to evaluate the hypothesis.

PRELIMINARY RESULTS--Preliminary results suggest the frequency ratio has an influence on metabolic efficiency. If the findings hold, this would imply that effective prosthesis selection depends on ambulation speed, user mass, and prosthetic foot/ankle stiffness and should be individualized. Furthermore, the frequency ratio may be used as a parameter in the design and manufacture of prosthetic foot/ankle systems.



Dudley S. Childress, PhD; Andrew H. Hansen, MS
Northwestern University Rehabilitation Engineering Program, 345 East Superior Street, Room 1441, Chicago, IL 60611; email: d-childress@nwu.edu; web: http://www.repoc.nwu.edu/

Sponsor: National Institute on Disability and Rehabilitation Research, Washington, DC 22202

PURPOSE--The roll-over shape of a prosthetic foot is defined as the sagittal plane geometry to which the foot deforms during the single-limb stance phase of walking. Recent research in our laboratory has shown that this roll-over shape is important to walking performance. Our studies suggest that prosthetic feet can be represented dynamically by their quasi-static properties. Two methods of measurement of roll-over shape, a quasi-static method and a direct method, have been developed in our laboratory.

METHODOLOGY--To measure the quasi-static roll-over shape, we load prosthetic feet, attached orthogonally to an instrumented pylon, mechanically at 5 orientations (60, 75, 90, 97, and 105°) in a prosthetic foot-loading apparatus. Center of pressure (COP), force, and deflection data are collected while gradually loading and unloading each foot. Although most feet show hysteresis in the loading/unloading cycle, only data associated with the loading part of the cycle are used.

  To measure the roll-over shape directly, we collect sagittal plane kinematic data of two markers, one at the fibular head (knee) and one at the lateral malleolus (ankle), to assess the forward progression component of the COP.

PROGRESS--Twelve prosthetic foot roll-over shapes have been measured using the quasi-static method. Six prosthetic and four human feet have been measured using the direct method.

RESULTS--For a given load condition, the location of the COP and the deflection at that load can be determined under the quasi-static roll-over shape method for all five foot orientations. Plotting deflection of the foot versus distance from the pylon to the COP gives the approximate shape each foot will take during stance phase roll-over under that particular load.

  The roll-over shape from direct measurement is created by transforming the COP movement in a "lab-based" coordinate system into its equivalent movement in a "shank-based" coordinate system. The ankle marker is the origin in the shank-based coordinate system, while the vector from ankle marker to knee marker determines one of its axes. The trajectory of the COP in shank-based coordinates determines the roll-over shape.

FUTURE PLANS--Clinical relevancy of roll-over shapes in prosthetics and orthotics lies in correct shapes and in correct placement of these shapes with respect to the lower limb. For example, prosthetic alignment procedures may serve to align the roll-over shapes of prosthetic feet toward favorable locations. A favorable shape may be suggested from ambulation studies of control subjects, using the direct method.

  We plan to measure more roll-over shapes of controls. Knowing rocker radii of able-bodied feet may lead to improved designs for prosthetic feet, ankle-foot orthoses, rocker shoes, and total contact casts.




Colin P.U. Stewart, MD, DMR; Jeremy R. Linskell, BSc, MSc; A. Kenneth MacDonald, HDipP&O, MSc
Tayside Rehabilitation Engineering Services, Dundee Teaching Hospitals NHS Trust, Dundee Limb Fitting Centre, Broughty Ferry, Dundee DD5 1AG, Scotland, UK; email: cstewart@dth.scot.nhs.uk

Sponsor: None listed

PURPOSE--Active persons with lower limb amputation have available to them an increasing number of commercial prosthetic systems and components. Seattle Limb Systems (formerly M+IND Corporation) has recently introduced two new devices to its range, namely the Voyager Foot, described by the manufacturer as being highly dynamic and offering improved stability, and the Select Pylon which is said to enhance the performance of Seattle feet. The purpose of this study is to evaluate these two devices, both clinically and biomechanically.

METHODOLOGY--Five subjects with established transtibial amputation, displaying moderate to high activity levels, have been selected from the local patient population. New prostheses with the capability of having the combination of their internal components (with the exception of the socket) changed, without affecting the outward appearance, have been fabricated and delivered to them in the normal manner. Thus, four different combinations of components are possible within each prosthesis, using the Voyager Foot, the Select Pylon, and the patient's regular shin-tube and foot. In this way, the gait analysis researcher involved in this study will be blinded to the componentry permutation being used by each subject at the time of the gait analysis tests.

  For each combination of components, and following a period of accustomization, each subject will return for gait analysis testing in the Dundee Gait Laboratory. A series of investigations will be performed, including standard 3-D gait analysis, force plate analysis at higher frequencies, shock-loading investigations using a shockmeter, and energy cost studies involving Physiological Cost Indices based on heart rate measurements.

  In addition to the acquisition of data as described above, each subject will be asked to complete, for each build configuration, an evaluation form for the purpose of recording his/her subjective opinion on the performance of the prosthesis.

  The gait data will be analyzed and interpreted by members of the Tayside Rehabilitation Engineering Services Clinical Gait Analysis Service.

PROGRESS--At the time of writing, all five subjects have had new prostheses fabricated and the data acquisition and analysis phase will be commencing in the near future. It is envisaged that the study will be completed by March 1999.

RESULTS--Although no results have been obtained yet, the following areas could yield useful information:

  1. Gait analysis will identify alterations in the gait pattern which may be interpreted in terms of the mechanical differences in the various configurations.
  2. Force plate analysis at higher frequencies and the shockmeter investigations will help to identify the load acceptance performance of each configuration and hence comparison between them.
  3. Physiological Cost Index measurements will assist in determining the relative efficiency of each combination of components.
  4. Subjective comparison by the users will determine whether or not they perceive any benefit from the new componentry under investigation.


Go to top.





Last revised Thu 04/29/1999