Previous |
Contents |
Next |
|---|
[030] EFFECT OF THE BANKART LESION ON ANTERIOR JOINT STABILITY WITH SIMULATED GLENOHUMERAL MUSCLE FORCES
Thay Q. Lee, MS; Patrick J. McMahon, MD
Orthopaedic Biomechanics Laboratory, PACT/Department of
PM&R, VA Medical Center, Long Beach, California, 90822;
Department of Orthopaedic Surgery, University of California,
Irvine, Irvine CA 92717; email:
tqlee@pop.long-beach.va.gov
Sponsor: Department of Veterans Affairs, VA
Rehabilitation Research and Development Service, Washington,
DC 20420
(Project #A880-RA)
METHODOLOGY--To study the effect of the Bankart lesion and joint effusion on glenohumeral joint kinematics and kinetics requires a building of a custom shoulder loading apparatus that utilizes a six axes load cell, six degrees of freedom (DOF) electromagnetic tracking device, four differential variable reluctance transducers (DVRT), and a video digitizing system (VDS). Using this device, we will biomechanically investigate both the passive and active stabilizing structures of the glenohumeral joint.
PROGRESS--The custom glenohumeral joint tensile testing jig, shoulder translation testing jig, and a shoulder loading apparatus equipped with six axes load cell and magnetic tracking devices have been completed. The translation testing jig permits the testing of passive restraints of the glenohumeral joint. The custom loading device permits the application of each individual muscle forces in the shoulder while permitting simultaneous measurement of the glenohumeral joint compression force and kinematics. The joint compression force is measured with the six axes load cell and the glenohumeral joint kinematics is measured with the magnetic tracking device. In addition, the inferior glenohumeral ligament (IGHL) origin as well as the permanent stretching of the IGHL tissue substance has also been determined.
PRELIMINARY RESULTS--Biomechanical results of the IGHL tensile tests showed that eight (66 percent) of the b-l-l-b complexes failed at the glenoid insertion site (representing the Bankart lesion), two (17 percent) at the humeral insertion site, and two (17 percent) at the anterior band midsubstance. Yield load was 579±63 N (mean±SEM) and the ultimate load was 676±59 N. Yield stress was 9.4±2.3 MPa and the ultimate stress was 10.8±2.3 MPa. Yield deformation was 11.9±0.5 mm and the ultimate deformation was 14.2±0.4 mm. On average, the permanent deformation, the difference between yield and ultimate deformation, was only 2.3 mm. Strain at failure was largest at the glenoid insertion of the anterior band on the labrum (17.1±1.7 percent). The humeral insertion strain (17.5±2.4 percent) was greater than at the anterior band midsubstance (8.9±1.0 percent). Interestingly, there was also considerable strain in the anterior band insertion on the glenoid proximal to the labrum (12.1±1.8 percent). Histology results showed that the anterior band of the IGHL has two attachments at the glenoid insertion site: the labrurn and the anterior glenoid neck. The labral attachment is comprised of poorly organized collagen fibers that have direct insertion into the fibrocartilagenous labrum. The glenoid attachment is composed of dense collagen fibers that extend proximal to the articular margin of the glenoid and have an indirect insertion into the bone at a distance of more than 10 mm. There is only slight permanent stretching of the anterior band with the initial traumatic dislocation. This experimental model demonstrates that most of the strain occurs at the insertion sites, with failure occurring most often at the glenoid insertion. Current operative techniques for the treatment of anterior glenohumeral instability that repair only the labral attachment of the anterior band of the IGHL are insufficient. With surgical repair of the of the Bankart lesion, the glenoid attachment should be incorporated into the labral attachment repair and only slight shortening of the anterior band (about 6 percent) is necessary to restore the joint anatomy after initial glenohumeral dislocation.
FUTURE PLANS--Fresh-frozen full upper limb specimens will be mounted by rigidly attaching the scapula on the shoulder testing apparatus. The muscles simulated are those active in abduction of the joint: deltoid (middle and upper portions), supraspinatus, subscapularis, and the infraspinatus/teres minor. The scapula will be positioned in 30° of abduction and muscle force will be applied to the tendons of the simulated muscles until the shoulder joint reaches 90° of abduction. The force in the infraspinatus/teres minor muscles will then be increased for maximum external rotation. Glenohumeral joint motion will be measured with a six DOF magnetic tracking device. Three DVRT's and continuous VDS will be used to determine in-situ strain in the IGHL. The IGHL will then be tested in tension to determine the in-situ stress. For the measurement of the resultant glenohumeral joint compression force, a six axes load cell will be used. A Bankart lesion will then be surgically created and the test repeated again. At the completion of joint motion, each glenohumeral joint will be dissected and the center of curvatures for the humeral head and the glenoid will be determined.
[031] BIOMECHANICS OF THE PATELLOFEMORAL JOINT AND PERIPATELLAR RETINACULUM
Thay Q. Lee, MS
Orthopaedic Biomechanics Laboratory, PACT/Department of
PM&R, VA Medical Center, Long Beach, California, 90822;
Department of Orthopaedic Surgery, University of California,
Irvine, Irvine CA 92717; email:
tqlee@pop.long-beach.va.gov
Sponsor: Department of Veterans Affairs, VA
Rehabilitation Research and Development Service, Washington,
DC 20420
(Project #A881-RA)
PURPOSE--The purpose of this study is to quantify the structural and mechanical properties as well as the in-situ stress and forces in the peripatellar retinaculum. This study is designed to quantitatively assess the functional role of the peripatellar retinaculum, in addition to determining the effects of lateral patellar retinaculum release, tibial rotation, and complete transverse patellar fracture (defect).
METHODOLOGY--Fresh frozen human cadaver knees will be used with an Instron machine and a custom patellofemoral joint loading jig that permits the measurement of strain in the peripatellar retinaculum while controlling the knee flexion angle, degree of tibial rotation, and tension in the quadriceps muscles (rectus femoris/vastus intermedius as well as vastus medialis and vastus lateralis). The patellofemoral contact pressures will be measured using Fuji pressure-sensitive film. The mean strain in the peripatellar retinaculum will be measured using six differential variable reluctance transducers (DVRT), and the surface strain will be measured using the video digitizing system (VDS). To determine the in-situ stress and forces in the peripatellar retinaculum, the in-situ strain is first measured in intact knees. The peripatellar retinaculum is then isolated for tensile testing to determine both the structural and mechanical properties. The results will then be correlated to compute in-situ stress and forces in peripatellar retinaculum.
PROGRESS--The patellofemoral joint loading jig has been built. This jig allows 5 degrees of freedom (DOF) at the femur and 3 DOF at the tibia. Muscle loading can be accomplished through stainless steel clamps, an adjustable pulley system, and lead weights. This custom jig permits independent application of all knee extensor muscle (rectus femoris/vastus intermedius, vastus lateralis, vastus medius, iliotibial band) forces in its anatomic/physiologic orientation at any knee flexion angle. This jig also permits simultaneous measurement of the patellofemoral joint contact pressures and kinematics using Fuji pressure sensitive film and a 3-D magnetic tracking device, respectively. We have also determined the anatomic/physiologic loading conditions for the in-vitro testing of the patellofemoral joint. This loading configuration was used for the quantification of the changes in the patellofemoral contact pressures as well as the in-situ strain in the peripatellar retinaculum resulting from tibial rotation. The tibial rotation had a significant effect on the patellofemoral contact pressures and the in-situ strain in the retinaculum. The increase in the patellofemoral contact pressures due to tibial rotation was observed for both internal and external rotation of the tibia with the effects being greater for the external tibial rotation. In addition, data show the greatest increases in the patellofemoral contact pressures occur at full extension. The results from this study suggest an inverse relationship between increasing knee flexion angles and the percentage change in mean patellofemoral contact pressures with rotation, especially external rotation. The results also suggests that there is a decrease in the strain of the contralateral side of the retinaculum with respect to the direction of the tibial rotation.
FUTURE PLANS--We plan to determine the tissue strain and stress in the peripatellar retinaculum as well as the patellofemoral contact pressures for lateral retinacular release and transverse patellar fracture (defect). Peripatellar retinaculum is an essential, complex, anisotropic, multilayered tissue for patellar stabilization. The residual strain and stress in these tissues, however, still needs to be quantified and further biomechanical and histological studies are necessary to accurately predict the in-situ stress in this complex tissue structure.
[032]FRACTURE OCCURRENCE AND HEALING IN SPINAL CORD INJURED PATIENTS
Beatrice Kiratli, PhD
VA Medical Center, Palo Alto, CA 94304
Sponsor: Department of Veterans Affairs, VA
Rehabilitation Research and Development Service, Washington,
DC 20420
(Project #A2014-RA)
No report was received for this issue.
[033] MECHANICAL LOADING EFFECTS ON CARTILAGE REPAIR AND REGENERATION
Robert Lane Smith, PhD; Dennis R. Carter, PhD; Andrew
Hoffman, MD; David J. Schurman, MD
Rehabilitation R&D Center, VA Palo Alto Health Care
System, Palo Alto, CA 94304; Dept of Medicine, Stanford
University School of Medicine, Stanford, CA 94305; Dept of
Functional Restoration, Stanford University School of
Medicine, Stanford, CA 94305; email:
smith@roses.stanford.edu
Sponsor: Department of Veterans Affairs, VA
Rehabilitation Research and Development Service, Washington,
DC 20420
(Project #A857-RA)
PURPOSE--This study addresses mechanisms by which mechanical loads influence rehabilitation of joint function through tissue repair and regeneration. The goal is to develop information applicable to interventional techniques for restoration of joint function in arthritis. The experimental aims seek to establish the effects of varying levels and frequencies of hydrostatic pressure on articular chondrocyte matrix synthesis, proteoglycan gene expression, and cartilage type-specific collagen synthesis. The hypothesis addressed in this study is that cartilage repair and regeneration can occur given proper mechanical and biological stimulation. The process of understanding the mechanical loading conditions that augment cartilage matrix synthesis will permit development of a functional implantable cartilage-like tissue-substitute for arthritic joints.
METHODOLOGY--The experiments involve the isolation and culture of normal bovine articular chondrocytes in high density primary monolayer cultures for exposure to mechanical loading regimens. The cells are collected by sterile dissection and enzymatic dissociation of cartilage from radiocarpal joints of adult bovines (age 5-7 yrs). The mechanical loading regimens include application of intermittent hydrostatic pressure at levels of 5, 10, 15, and 20 MPa at frequencies of 0.1, 1, 5, and 10 Hz. The mechanical loading is carried out by placing culture dishes containing the cells in high density monolayers into medium-filled bags contained in a high pressure vessel. After loading, the cell culture medium is collected for analysis of protein, proteoglycans, and glycosaminoglycan content. The cell monolayer is then extracted, using a modification of the guanidinium isothiocyanate methodology for recovery of total cellular nucleic acid. Gene expression is analyzed by semi-quantitative RT-PCR methodology, using cDNA and primer pairs designed for amplification of cartilage-specific markers. Quantification of mRNA signal levels is based on detection of 32P radioactivity. Replicate samples are tested for statistically significant differences using analysis of variance with multiple comparison testing by the general linear method applying Tukey's correction.
PROGRESS--The initial experiments concentrated on the cartilage type-specific collagen mRNA level, since collagen provides the framework for the extracellular matrix. Progress has involved optimization of the PCR conditions for analysis of type II and type I collagen mRNA levels relative to a reference signal for beta-actin. Use of a single reaction mixture for quantification of each collagen mRNA signal proved unsuccessful. Competition between primer species for type II collagen and beta-actin necessitated individual reaction mixes and required optimization of the individual primer sets as independent series. This work has progressed well. Five sets of chondrocyte cultures have been exposed to intermittent hydrostatic pressure at 10 MPa with a frequency of 1 Hz for periods of 2, 4, 8, 12, and 24 hrs and have yielded a profile for collagen expression.
RESULTS--The data obtained represented articular chondrocytes placed in monolayer culture in the presence of 10 percent serum and then mechanically loaded in serum-free medium following a 24-hr period of serum-washout. Under these conditions, type II collagen expression was relatively abundant and the levels of mRNA did not vary significantly during the periods tested. Detection of type I collagen expression required an increased number of cycles of PCR and the existence of type I collagen signal varied with cell preparation. Signal levels for beta-actin mRNA revealed that this marker of chondrocyte metabolism was also relatively abundant and that the levels of expression did not vary with exposure to intermittent hydrostatic pressure.
FUTURE PLANS--The experimental protocols will continue by examining effects of different levels of intermittent hydrostatic pressure on the cells. Quantification of the chondrocyte response will be expanded to include analysis of aggrecan core protein mRNA signal. In addition, effects of hydrostatic pressure on protein and glycosaminoglycan levels in the culture medium will be determined.
[034] THE EFFECT OF HYDROSTATIC PRESSURE ON INTERVERTEBRAL DISC METABOLISM
Scott D. Boden, MD; William C. Hutton, DSc; William A.
Elmer, PhD; Steve Hyon, BS; Yasumitsu Toribatake, MD;
Katsuro Tomita, MD; Gregory A. Hair, PhD
Emory University, Department of Orthopaedics, Decatur,
GA 30033; email: william_hutton@emory.org
Sponsor: Department of Veterans Affairs, VA
Rehabilitation Research and Development Service, Washington,
DC 20420
(Project #A834-3RA)
PURPOSE--The etiology of many causes of low back pain can be linked to disc degeneration, a condition that may be the first step in the development of osteoarthritis and stenosis in the lumbar spine. It is commonly held that the forces applied to the intervertebral disc play a major role in causing disc degeneration. However, the nature of the association between applied force and disc degeneration remains obscure. It has been suggested that in some cases the applied force generates a mechanical stress that acts in cyclic fatigue to damage the disc at a rate faster than the normal biologic repair mechanism. This is a popular and a convenient explanation, but fails to explain the underlying biologic processes and the fact that intervertebral disc degeneration can also result from insufficient loading such as that seen with prolonged space travel and bed rest.
This project tests the hypothesis that hydrostatic pressure directly affects the synthesis of collagen and aggrecan (the cartilage-specific proteoglycan) by intervertebral disc cells. The long-term objective is to determine how hydrostatic pressure is transduced into the biological response within the intervertebral disc cells. Specifically, the experiments described are designed to determine how hydrostatic pressure influences the mechanisms responsible for normal disc maintenance and to identify specific biomarkers that can be used to indicate deviations from normal values. Knowing which steps in disc cell metabolism are sensitive to hydrostatic pressure could become an important asset in understanding and preventing disc degeneration.
METHODOLOGY--We have been carrying out experiments using canine intervertebral disc cells cultured in alginate and subjected to 1 MPa pressure for 9 days. The culture system takes advantage of an alginate (a linear copolymer of D-mannuronic acid and L- guluronic acid) which polymerizes to form a gel in the presence of calcium ions. This system has been the choice of a number of investigators in the study of cartilage cell metabolism. In our experiments, the intervertebral disc cells are exposed to a specific value of hydrostatic pressure inside a pressure chamber. The cells are then recovered by dissolution of the alginate gel in calcium-free buffer containing the chelating agent, ethylenediamine tetracetate (EDTA). The biosynthesis of the extracellular matrix (ECM) components (the proteoglycans and collagen) are then measured.
PROGRESS--Using a specially designed pressure vessel, results show a differential response of nucleus and anulus cells to 1 MPa of pressure (0.1 MPa;sgatmospheric pressure). Incorporation of [3H]-proline and [35S]-sulfate into collagen and aggrecan, respectively was stimulated in nucleus cells and inhibited in anulus cells compared to controls at 0.1 MPa.
FUTURE PLANS--We will continue these experiments at different values of hydrostatic pressure in the range of 0.1 MPa to 2.0 MPa (at least two values above and two below 1 MPa). We will also determine whether the biosynthetic response is a reflection of changes in mRNA levels for collagen types I and II and the aggrecan core protein using the RT-PCR methodology. Using the same culture system and the same biochemical and molecular methodologies, we will measure the biosynthetic response of intervertebral disc cells to hydrostatic pressure applied cyclically. To carry out these experiments, we have designed and built a cyclic pressure apparatus that has been tested and observed to be fully functional.
[035] BIOMECHANICS OF FOOT DEFORMITIES AND ALTERNATIVES FOR SURGICAL CORRECTION
Bruce J. Sangeorzan, MD
Department of Orthopedics, Harborview Medical Center
Orthopaedic Biomechanics Lab, Seattle, WA 98104; Section
Orthopedics, Puget Sound VA Medical Center, Seattle, WA
98108-1597; email: BSANGEOR@U.washington.edu
Sponsor: Department of Veterans Affairs, VA
Rehabilitation Research and Development Service, Washington,
DC 20420
(Project #A2034-RA)
No report was received for this issue.
[036] DEVELOPMENT OF A SHOULDER AND ELBOW MUSCULOSKELETAL MODEL: CLINICAL AND ERGONOMIC APPLICATIONS
F.C.T. van der Helm; A.L. van Lunteren; L.A. Rozendaal;
J.H. de Groot; S.H. Stroeve; B.L. Kaptein, G.G. Brouwn
Man-Machine Systems and Control Group, Department of
Mechanical Engineering, Delft University of Technology,
Delft, The Netherlands; email:
F.C.T.vanderHelm@wbmt.tudelft.nl
Sponsor: Dutch Organization for Scientific Research (NWO)
PURPOSE--In the past years, a detailed musculoskeletal model of the shoulder has been developed, including the scapular and clavicular motion. The purpose of our research is to obtain a better knowledge of the biomechanics and control of the upper limbs, in order to improve the clinical diagnosis and treatment, and ergonomic prevention of disorders.
METHODOLOGY--We have extended the existing shoulder model with the flexion/extension joint and the pro/supination joint. The current model consists of 31 muscles and muscle parts, represented by 115 muscle lines of action, wrapped around muscle contours when applicable. In a separate morphological study, the sarcomere length of all shoulder and elbow muscles has been measured using laser diffraction. From these data, the optimum fiber length, pennation angle, and tendon length can be derived, and hence the force-length curve. Image processing techniques are being used to assess the muscle attachments in vivo from MRI and CT-scans.
Rozendaal has analyzed the role of geometrical factors (change of moment arms), intrinsic muscle properties (stiffness, viscosity), and reflexive muscle properties (length and velocity feedback, force feedback, including time-delays), on the impedance of a linearized musculoskeletal model. Using a robotic manipulator, model parameters for the intrinsic and reflexive properties are identified. Stroeve analyzed the properties of a combined feedforward and feedback (including time-delays) controller using a neural network.
In collaboration with the Department of Orthopaedics, University Hospital Leiden, a Flock-of-Birds system has been adapted and calibrated for motion recording of the shoulder (including scapula and clavicle) and elbow. Recorded motions can be directly used for a inverse dynamic analysis with the musculoskeletal model. A case study of a brachial plexus subject showed the potential of the model in predicting the mechanics of a disabled shoulder as well as the postoperative results.
In collaboration with the Department of Human Movement Sciences, Vrije Universiteit Amsterdam, recordings have been made of motions, external forces, EMG, and oxygen uptake during wheelchair propulsion. These data are used for assessment of mechanical load (e.g., of the glenohumeral joint) and of the power generation of the shoulder and elbow muscles in order to analyze the cause of energy losses resulting in a low mechanical efficiency.
PROGRESS--The shoulder model has been improved with better parameters (optimum muscle length), an elbow extension and control loops with muscle spindles and Golgi tendon organs. Rozendaal has shown that it was feasible to calculate the optimal feedback gains for individual muscles. By force perturbations of a robotic manipulator, the role of muscle spindles could be identified from the EMG. Simulations with neural networks showed that there was a high interaction between the feedforward (planning) and feedback control. The model has become reasonably user-friendly, and hence applicable in clinical and ergonomic settings.
[037] DETECTION OF OSTEOPOROSIS THROUGH COMPUTERIZED ANALYSIS OF DISTAL RADIUS X-RAYS
Carlos A. Wigderowitz, MSc, MD; Eric W. Abel, PhD; David
I. Rowley, MD
University of Dundee, Dundee Royal Infirmary, Department
of Orthopaedics, Dundee DD1 9ND, Scotland, UK; email:
cawigder@btinternet.com
Sponsor: University of Dundee, Dundee DD1 9ND, Scotland, UK
PURPOSE--The purpose of this study is to develop a sensitive and inexpensive method to diagnose osteoporosis based on computerized analysis of digitized x-rays. The quantification of structural rather than densitometric properties is the aim of this project. The distal radius has been chosen so far because it usually develops early osteoporotic signs and is an area prone to osteoporotic fractures. Further the cancellous bone content is high and radiographs of large numbers of patients are readily available.
METHODOLOGY--Distal radius x-rays are digitized using a high-resolution flat-bed scanner, interfaced to a personal computer. The images are converted to grey-level matrices and the grey-level variations are analyzed using frequency domain techniques, based on the Fast Fourier Transform (FFT). The magnitude peaks in the FFT are identified and their directions marked, permitting the determination of the preferred trabecular orientation in space. The magnitude of the peaks, expressed as a fraction of the total magnitude of the transform, permits a quantification of the trabeculae, expressed as a textural property of the image.
PROGRESS--The method is being considered for utilization in osteoporosis screening in places where densitometric methods are not available and where x-ray images have to be sent through telephone connections to be analyzed. Biomechanical studies are on their way to determine the ability of the method to predict the mechanical strength of bones.
RESULTS--Images have been analyzed in this way for a large number of clinical x-rays, and the results suggest that the ability to detect osteoporosis is similar to that of densitometric methods such as dual and single photon absorptiometry. The method is also able to detect the alterations that occur with fractures of the bones. The preliminary results with biomechanical testing show that the structural parameters have a significant correlation with the mechanical strength of the bones tested.
FUTURE PLANS--Comparative studies are being planned between spectral analysis and densitometric methods both on clinical and in vitro grounds. Other structural parameters, such as bone size and shape, are being considered, and combinations of these factors are being studied as bone strength predictors.
[038] MICROCRACK MORPHOLOGY IN FATIGUED COMPACT BONE
T. Clive Lee, MD, PhD, FRCSI; Fergal J. O'Brien, BA,
BAI; Felice Pellegrini; David Taylor, PhD
Department of Anatomy, Royal College of Surgeons in
Ireland, Dublin 2; Politecnico di Torino, Turin, Italy and
Mechanical Engineering Department, Trinity College, Dublin
2, Ireland; email: tclee@rcsi.ie
Sponsor: Health Research Board of Ireland, Royal College of Surgeons in Ireland
PURPOSE--Bone remodelling repairs fatigue-induced microdamage. Deficient repair results in fragility fractures, yet little is known of the behaviour of fatigue microcracks in bone. We have identified a series of fluorescent stains which are as effective as the standard method--basic fuchsin--in detecting microcracks but are also site-specific as they chelate Ca2+ ions lining the crack walls. We have shown that these chelating agents can be used sequentially to differentiate artefactual from test-induced damage and to label crack propagation. This study aims to refine this detection method, as binding strengths and levels of fluorescence vary among the agents, to use it to characterise microcrack initiation, propagation and coalescence in compact bone, and to relate microcrack generation to changes in the material properties of bone.
METHODOLOGY--Using microcracks generated during tensile testing, the minimum concentration of each chelating agent required to label the crack, as seen using epifluorescence, was established. Using these minimum concentrations, the growing microcracks were labelled at intervals with different agents to establish the optimal dye sequence.
PROGRESS--Coupons of bovine compact bone (35×25×2.5 mm) were prepared, a notch cut in one side and tested in tension to generate microcracks. Notches were cut obliquely with a diamond saw to avoid brittle fracture. The concentration of individual chelating agents applied (alizarin complexone, calcein, calcein blue, oxytetracycline, and xylenol orange) was varied from 0.001 to 1 percent to establish the minimum effective concentration necessary to bind to the walls and so label the crack, visualized using epifluorescence microscopy. Using these dosages, tensile tests were stopped at intervals and chelating agents applied to establish the optimal dye sequence for labelling segments of the propagating microcrack. Beams of compact bone (60×10×3 mm) were machined, polished, and fatigue tested in 3 or 4 point bending to 7,000 microstrain at 10 Hz. The optimal concentration and sequence of chelating agents were applied and growth of fatigue-induced microcracks monitored.
PRELIMINARY RESULTS--The optimal concentrations of each agent were: alizarin=0.015 percent, calcein=0.005 percent, calcein blue=0.015 percent, oxytetracycline=0.015 percent, xylenol orange=0.045 percent. The best single dye was found to be xylenol orange in a two-dye sequence (xylenol and calcein) and a three-dye sequence (alizarin, xylenol, and calcein). However, where a dye sequence was used, there was some evidence of the second dye displacing the first from the crack wall, suggesting either a mass effect or a greater affinity for exposed Ca2+ ions. Microcracks were easily generated in 3 point bending but were widely dispersed. Four point bending concentrated the stresses over a smaller area and the location and generation of microcracks was more easily predicted. Growth of fatigue-induced microcracks was labelled with a sequence of three chelating agents.
FUTURE PLANS--The stoichiometry of the chelating agents will be further refined by means of potentiometric titrations and X-ray crystallography. This method will then be applied to compact bone fatigue tested in four point bending and microcrack number and growth in length will be related to loss of bone stiffness and strength, for the first time. Microcrack morphology in serial sections will be recorded and, via a new computer package, a 3-D reconstruction made. This would be the first such attempt to show the passage of a crack through bone in 3-D. These data will be used to calibrate a finite element model of bone fatigue. Such a computer model, based on theoretical prediction and supported by experimentally derived material properties and morphological data would be a powerful tool in fracture prediction and orthopaedic prosthesis design.
[039] QUANTITATIVE FUNCTIONAL ANATOMY OF THE UPPER EXTREMITY
DirkJan H.E.J. Veeger, PhD; F.C.T. Van der Helm; K.N.
An; L.A. Rozendaal
Faculty of Human Movement Sciences, Vrije Universiteit,
1081 BT Amsterdam, The Netherlands; Department of
Measurement and Control, TUD, 2628 CD Delft, The
Netherlands; Orthopaedic Biomechanics Lab., Mayo Clinic,
Rochester MN 55905; email: h_e_j_veeger@fbw@vu.nl
Sponsor: Netherlands Organisation for Scientic Research (NWO-MW).
PURPOSE--Quantitative data on the morphology of shoulder and arm are needed as a basis for analysis of the load on the shoulder and arm, based on arm movement registration in wheelchair propulsion, activities of daily living, and vocational activities; analysis of the outcome of shoulder arthrodeses; and interpretation of in-vivo human palpation data.
METHODOLOGY--We are building a collection of musculoskeletal parameters of the shoulder and arm for biomechanical modelling of that region. Many of the clinical and ergonomical problems of the shoulder are the result of the complex co-ordination of the muscles involved in the control of shoulder movements and joint stabilization. This complexity of the human shoulder and arm cannot directly be studied, but needs the application of a 3-D biomechanical model. In addition, the 3-D nature of shoulder and arm motion and the covert motions of the scapula, require a highly sophisticated 3-D movement analysis.
PROGRESS--To date, morphological parameters have been collected on the shoulder mechanism as well as the arm. These data comprise 3-D insertion sites of muscles, the 3-D orientations of axes of rotation for elbow flexion/extension and pro/supination, and the rotation center of the glenohumeral joint. Also collected were data on muscle morphology such as muscle mass or physiological cross sectional area. These data have been implemented in a biomechanical model of the shoulder and arm. With the use of the model it is possible to calculate muscle forces, tensions in ligaments, and reaction forces in joints of the shoulder. Parallel to the model development, 3-D techniques for the measurement of kinematics have been developed. The program has been proven to be useful in the development of a sophisticated 3-D model of the shoulder mechanism that has been applied in the prediction of optimal fusion angles of shoulder arthrodeses after injury of the brachial plexus, in the analysis of the positioning of endoprostheses and operation techniques used, and in the quantification of mechanical load on the shoulder joint in manual wheelchair propulsion. Currently, the model is being applied in the analysis of the effect of glenohumeral joint stabilization requirements on the mechnical and physiological load of the manual wheelchair propulsion. Morphological data on these research projects are now available at http://www-mr.wbmt.tudelft.nl/shouder/dsg/dsg.info.html.
[040] OBSTACLE AVOIDANCE TRAINING WITH COMPUTER SIMULATED ENVIRONMENTS: A PILOT STUDY
David A. Brown, PhD, PT; David L. Jaffe, MS
Center for Disease Prevention, Stanford University,
Stanford, CA 94305; email: jaffe@roses.stanford.edu
Sponsor: Department of Veterans Affairs, VA
Rehabilitation Research and Development Service, Washington,
DC 20420
(Pilot Project #A1848-PA)
PURPOSE--Our work has focused on measuring characteristics of stepping-over responses (SORs) in overground walking and with computer-simulated objects on a treadmill. We have accomplished both of our stated objectives: to develop a method for displaying computer-generated obstacles so as to elicit stepping-over responses on a treadmill, and to demonstrate that these computer-generated SORs showed kinematic characteristics that were similar to overground SORs. In addition, we have conducted preliminary tests designed to show that our proposed training regimen (e.g., using computer-generated obstacles on a treadmill) is effective in improving overground SORs.
METHODOLOGY--We employed a head-mounted display to show a composite video image of the subjects with stationary virtual objects at their feet. They were asked to continually step over the obstacles with their left foot while on the treadmill.
PROGRESS--Five subjects have, so far, been enrolled in this part of the pilot project. All subjects are young, nonimpaired controls. Each person participated in two stepping sessions. In the first session, subjects were asked to walk down a walkway at a fixed speed and step over rectangular boxes of varying height and length. In the second experiment, subjects were asked to walk on a motorized treadmill while viewing their own legs and a computer-generated obstacle. The subjects were told to continually step over these simulated obstacles. Kinematics were measured using sagittal plane motion analysis. We measured hip, knee, and ankle angles, as well as foot height, step length, and step time when the leg was passing over the object.
The data collected show that we were successful in producing a method for eliciting SORs using computer-generated obstacles: they generate kinematic trajectories that are similar to overground SORs.
The second experiment involved three elderly subjects, one nonimpaired and two with stroke more than 3 years prior to testing. Both stroke subjects had regained fully independent ambulation but showed a noticeable lack in step height and stride length, which is exacerbated when asked to step over obstacles. Each person was initially tested on a 30.5 m course with nine foam obstacles 3 m apart. They walked this course as fast as possible, while carefully trying to avoid colliding with the objects. Then they practiced the proposed training regimen with simulated objects over a 3-day period, and were retested on the obstacle course.
RESULTS--Our work has led to the following conclusions:
[041] DESIGNING A BALLISTIC WALKING BIPED: ACTIVELY VARIABLE PASSIVE STIFFNESS
Richard Q. van der Linde, MSc; Anton van Lunteren, PhD;
Dick H. Plettenburg, MSc
Wilmer Group, Department of Mechanical Engineering,
Delft University of Technology, Delft, The Netherlands;
email: r.q.vanderlinde@wbmt.tudelft.nl
Sponsor: Delft University of Technology; Delft, The Netherlands; University of Twente, Enschede, The Netherlands
PURPOSE--Previous research has shown that passive (or ballistic) walking is an energy-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 which 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. Bifurcation theory 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 optimized 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, and these need to be controlled.
PROGRESS--A mathematical model of a simple bipedal walking movement is built gradually. Each effect of a model extension is to be studied in order to expose its effect on system behavior. Simultaneously, experiments are performed to verify theoretical validity. These actions build a toolbox that is to result in a method for synthesis of a desired ballistic walking movement.
PRELIMINARY RESULTS--A working prototype biped already has been constructed, and is capable of stiff-legged, stable, three dimensional walking on a horizontal surface. The ballistic walking motion is actuated solely by a periodic push-off, active for only one-forth of a step cycle. Also an actively variable passive stiffness joint has been developed. It facilitates the combinination of the advantages of passive motion with active control. It is believed that this joint can improve step width as well as robustness for external disturbances of the prototype.
[042] THE DEVELOPMENT OF A DIRECT ULTRASOUND RANGING SYSTEM FOR THE QUANTIFICATION OF AMBULATION
Dudley S. Childress, PhD; Richard F. ff. Weir
Northwestern University Rehabilitation Engineering
Research Program, Chicago, IL 60611; email:
d-childress@nwu.edu;
Web: http://www.repoc.nwu.edu/
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 22202
PURPOSE--Full gait analysis capabilities for comprehensive studies of walking tend to be costly to develop and maintain. It was hoped that a simpler monitoring system could be developed which would capture some of the important information about gait performance but which would be relatively inexpensive to obtain, simple to use, and easy to maintain within small clinical offices. We developed the direct ultrasound ranging system (DURS) to meet these criteria.
METHODOLOGY--The DURS measures the instantaneous horizontal velocity of the approximate body center of mass (CM) during ambulation. The system accurately measures the periodic fluctuation in the forward velocity of the body trunk that results from the rising and falling of the CM during normal gait. DURS operates by emitting high intensity infrared pulses from a base unit at a rate of 22 pulses/s. A transponder worn by the subject is triggered by each infrared pulse to emit an ultrasound pulse. The time taken by the ultrasound pulse to travel from transponder to base unit is then measured. By calibrating for the speed of sound in air, this time difference can then be converted into a measure of the distance between the transponder and base unit. These distance samples are then stored and processed by a computer to yield the instantaneous velocity. Preliminary results suggest that the DURS can be utilized to accurately quantify certain parameters of ambulation. A patent application for our system has just been filed with the US Patent Office.
PROGRESS--The DURS is now in its second generation. The transmission and reception of both the infrared light and ultrasound signal have been improved to increase system range and robustness. This was accomplished by changing the configuration of the ultrasound and infrared emitters and receivers to enable the use of high-intensity infrared flooding. In addition, circuit changes and software upgrades were made. The DURS can now measure out to a distance of 14 m with standard deviations of less than 1 mm over the first 10 m, rising to 5 mm at 15 m.
FUTURE PLANS--The next stage of this project is clinical evaluation of six systems. In addition to gathering information about its use, this will also allow us to build up a database of pathological gait velocigrams. The DURS, when used in conjunction with a home video camera, offers an affordable and effective means of monitoring the recuperation of patients who have undergone corrective gait surgery, joint implant surgery, or who have been recently fitted with a lower-limb prosthesis. This device is portable, easy to setup and use, and provides results in real time, enabling changes to be tried and tested immediately.
[043] CHARACTERISTICS OF AMPUTEE GAIT
Dudley S. Childress, PhD; Stephanie B. Michaud, MS
Northwestern University Rehabilitation Engineering
Research Program, Chicago, IL 60611; email:
d-childress@nwu.edu;
Web: http://www.repoc.nwu.edu/
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 22202
PURPOSE--Persons with unilateral transtibial or transfemoral amputations can often, with proper prosthetic fitting and rehabilitation, walk proficiently with little or no gross gait deviation. However, even adept prosthetic walkers are still known to ambulate with a slower speed and greater energy expenditure than their nonimpaired counterparts. While some deviation from normal walking could be inevitable, it may be possible to make aided ambulation on prostheses more efficient and comfortable by refining the prosthetic systems presently used. A better understanding of prosthetic ambulation will potentially lead to improved designs in prosthetic componentry.
The purpose of the present study was to enrich the understanding of prosthetic walking by gathering data concerning the gait of persons with transtibial and transfemoral amputation, noting patterns consistent among each group, and making comparisons to the gait data of velocity-matched controls. Focus in this study was placed on gait parameters in the coronal plane; these parameters have been largely ignored in the literature, yet have potential significance for persons with lower limb amputation in terms of balance, stability, and comfort. In order to assess whether alterations in the coronal-plane gait parameters are observable in prosthetic walking, several of these parameters were measured using gait analysis. Deviations from control walking were noted.
METHODOLOGY--Six persons with transtibial, and three with transfemoral, amputation participated in the study. Kinematic data were acquired using a CODA 3 Motion Measurement System as subjects walked along a 7-m walkway; ground reaction forces were measured simultaneously with two AMTI forceplates embedded in the walkway. Gait parameters investigated include velocity, cadence, step and stride length, stance time, trajectories and velocities of the pelvis, rotations of the pelvis in the coronal and transverse planes, the three components of the ground reaction force, and the transverse moment generated on the forceplate. Data from three male controls previously obtained in the laboratory were used as a reference in the present study.
RESULTS--The results of this study emphasize the importance of examining gait parameters with respect to the speed of ambulation when making comparisons between different walkers. Also highlighted is the marked deviation between subjects and controls with respect to the parameters of gait in the coronal plane. Specifically, the following observations were considered to be significant:
FUTURE PLANS--This preliminary investigation characterized several aspects of prosthetic walking that have not previously been reported in the gait literature. This work suggests that there are several areas in which further study could lead to significant improvements in prosthetics practice. For example, the pattern of pelvic obliquity that was observed in the transtibial subject using a vertical shock pylon suggests further investigations are needed to determine the effect of vertical compliance mechanisms on the gait of persons with unilateral amputation. Also, deviant patterns of pelvic obliquity and an altered temporal relationship between the peaks of pelvic obliquity and the gait cycle observed in the transfemoral subject wearing a quadrilateral socket indicate that the effect of socket design on pelvic motion and other aspects of unilateral gait needs to be investigated.
We feel the characterization of prosthetic gait and the identification of factors inhibiting persons with amputation from walking more proficiently are important first steps toward the improvement of prosthetic design, fitting, and gait training.
[044] THREE-DIMENSIONAL SENSITIVITY ANALYSIS OF SWING PHASE TOE CLEARANCE DURING GAIT
Dudley S. Childress, PhD; Steven A. Gard, PhD
Northwestern University Rehabilitation Engineering
Research Program, Chicago, IL 60611; email:
d-childress@nwu.edu;
Web: http://www.repoc.nwu.edu/
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 22202
PURPOSE--Foot clearance during the swing phase of walking is determined by the segment lengths and joint angles of the swing limb, stance limb, and pelvis. Foot clearance, also known as floor clearance, is defined as the perpendicular distance between the plane of the walking surface and a particular point on the swing-leg foot. Normal walking requires that the effective length of the swinging limb be shorter than the distance between the swing-leg hip and the ground in order to achieve foot clearance. Effective leg-shortening allows the swinging limb to be advanced in front of the body without contacting the ground and interrupting the gait cycle. We developed methods for determining the influence of the stance- and swing-limb joints on foot clearance and effective leg length during ambulation. The sensitivity analyses indicate how incremental changes in hip flexion/extension and ab/adduction, knee flexion/extension, and ankle dorsi/plantarflexion of both the swing and stance legs can affect the swing-leg toe clearance.
METHODOLOGY--Kinematic data describing the instantaneous positions of the joint centers of rotation (CoRs) from the swing leg, stance leg, and pelvis, and the position of the swing leg's great toe were acquired on two subjects. One of the subjects was a normal ambulator, and the other was a subject with a known leg length discrepancy resulting from congenital abnormalities.
PROGRESS--The original sensitivity analyses that we developed--the toe clearance sensitivity (TCS) analysis and the hip-toe distance sensitivity (HTDS) analysis--were two-dimensional and limited only to analysis of the swing leg joints. Recently, we have become quite interested in the contributions of the stance leg and pelvis in achieving and affecting swing leg toe clearance. Therefore, the original TCS and HTDS analyses have been expanded to include the effects of the pelvis and stance leg joints on the swing leg toe clearance, based upon the instantaneous positions of the joints' CoRs, angular positions, and axes of rotation. These analyses indicate how rotations of the joints of the swing leg, stance leg, and pelvis affect swing leg toe clearance and hip-toe distance throughout the gait cycle.
RESULTS--We have yet to analyze all of our data, but preliminary data from normal walking shows that the stance leg ankle flexion/extension may have a much more significant effect on swing leg toe clearance than the literature has suggested. Other joint motions of the stance leg that appear to have a larger influence on swing leg toe clearance than we expected are knee flexion/extension, hip ab/adduction, and ankle inversion/eversion. Typically, toe clearance problems are attributed to the joints of the swinging limb during gait, but our results indicate that the stance leg and pelvis make significant contributions to foot clearance that cannot be ignored.
FUTURE PLANS--The TCS and HTDS analyses may prove to be beneficial for analyzing and diagnosing foot clearance problems in many different types of pathological gaits. Joints that have apparently normal kinematics during gait may still cause significant foot clearance problems if other joints in the kinematic chain (stance leg, pelvis, swing leg) have altered rotation patterns. The sensitivity analyses may also be useful for analyzing individual prosthetic components and assembled lower-limb prosthetic and orthotic appliances. We plan to work with orthopedic clinicians in developing practical applications and interpretation of the TCS and HTDS analyses.
[045] EXAMINATION OF ACTIVITIES AROUND STANDING
Dudley S. Childress, PhD; Janet H. Jhoun, PhD
Northwestern University Rehabilitation Engineering
Research Program, Chicago, IL 60611; email:
d-childress@nwu.edu;
Web: http://www.repoc.nwu.edu/
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 22202
PURPOSE--Standing upright is a mechanical feat, since the structure is comprised of a high center of mass with a small base of support; yet humans stand with minimal muscular effort. The question of how humans make the transition from standing to gait is addressed in this project. We have examined weight transfer from one limb to the other, walking in place, and gait initiation in forward, backward, and sideways directions. In addition to discovering insights into human locomotion, we hope to apply this knowledge to improve performance of persons who have difficulty initiating gait, such as those with bilateral lower limb amputation (TFA) and Parkinson's disease. TFAs have trouble starting gait, and many exhibit stop-and-go gait. Bilateral TFAs who wear "stubbies" usually display good standing abilities, but have a "fall-stop" type of gait. Each step is like an initial step. Training of persons who have walking difficulties is often based on concepts derived from steady-state walking. We believe the transitional stage of locomotion may reveal underlying mechanisms that make ambulation difficult for some people.
METHODOLOGY--We examined transitional periods of locomotion in six nonimpaired young adults who have no known balance or musculoskeletal disorders in weight transfer from one limb to the other, walking in place, and gait initiation in the forward, backward, and sideways directions. Pelvic motion and ground reaction forces and center of pressure were measured by the CODA-3 Motion Analysis System and two AMTI force plates. A minimum of three trials for each subject on each activity was recorded and analyzed.
PROGRESS--Data collection and analysis have been completed for all six subjects and mathematical models formulated. The first endeavor was a forward simulation of the period from initial limb toe-off to heel-contact. This simple two-dimensional model was composed of two links representing the trunk and leg segments in the sagittal plane. It also incorporates a rocker-based structure. This model, however, did not represent gait initiation well. Another model that incorporates an additional link to represent the pelvis has been proposed.
RESULTS--Similar results were found across the subjects as well as across the tasks, including vertical ground reaction forces, initial pelvic obliquity, and center of pressure excursions. Gait initiation studies in the literature have not yet addressed the role of the pelvis during the transition from steady-state standing to steady-state walking. We found that measurements of pelvic obliquity and vertical displacements of the markers placed on the right and left sides of the pelvis during the first step are markedly different from those found during the subsequent steps. The step-off side is elevated relative to the initial stance side. This rise is mainly due to hip hiking on the step-off side, probably in order to achieve surface clearance.
FUTURE PLANS--We hope to extend our experimental and modeling to better understand the dynamics of gait initiation.
[046] A SIMPLE MODEL FOR STANDING AND BALANCE
Dudley S. Childress, PhD; Richmond B. Chan, PhD; Janet
H. Jhoun, PhD
Northwestern University Rehabilitation Engineering
Research Program, Chicago, IL 60611; email:
d-childress@nwu.edu;
Web: http://www.repoc.nwu.edu/
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 22202
PURPOSE--Standing is a dynamic process involving balancing the body in a vertical position. In our efforts to understand balance involved in normal and pathological gait, we investigated the process of balance during standing. During standing, the center of mass (COM) of the body and the center of pressure (COP) move in what appears to be a random manner. Investigation of these motions may help in the understanding of standing balance. We have developed a new model describing human standing as a kind of tracking mechanism. This descriptive model will hopefully provide obvious indication of balance performance and suggest treatment for the balance-impaired, such as the elderly.
METHODOLOGY--A model describing human standing was developed by combining McRuer's control strategy for human tracking with Winter's model for standing based on the structure of an upright body. The resulting model indicates that during the process of standing, the velocity of the center of pressure (VCOP) is proportional to the acceleration of the center of mass (ACOM) with a time delay.
Three sets of experiments were used to support the validity of our model. Six nonimpaired controls between 20 and 40 years of age participated in one part of the research. A CODA-3 motion analysis system and two AMTI force plates were used to collect 30 s of position and force plate data while each subject stood quietly. Additionally, three subjects performed the following tasks on one force plate: standing on one foot (flat), and standing with raised heels of both feet (on the forefoot). In addition to the standing experimental protocol, two subjects were asked to balance a yardstick in the palm of their hand.
PROGRESS--Data collection and preprocessing have been completed, and initial analysis suggests that during the process of standing, the VCOP is proportional to the ACOM with a time delay, as our model had predicted. The stick balancing activity can also be described by the model. Agreement between the data and the prediction of the model suggests a similarity between human standing balance and human tracking behavior. In standing, the position of the COP tracks the position of COM. This behavior is a tracking behavior that is found in balancing a stick in the palm of one hand. Investigation on how the model applies to standing balance with increasing level of difficulty (i.e., standing with raised heels, or with one foot) is in progress.
RESULTS--The experimental data show that in general, the VCOP follows proportionately the curve of ACOM with a time delay, as predicted by our model. A coherence function was used to determine the extent to which ACOM is related to VCOP. A coherence of 70 to 90 percent was found on the subjects studied, which suggests agreement with the model. The model revealed a time delay associated between ACOM and VCOP. Autocorrelation analysis is used to compute this delay. Initial estimates for the time delay during the activities investigated fall within the 200 to 450 ms range.
The model has revealed a control strategy and a delay associated with normal standing. Outcomes of this project include possible new approaches for both the clinical screening and evaluation of persons for balance deficiencies, and the evaluation of the balance of prosthesis and orthosis users, which is key to good ambulation.
FUTURE PLANS--We plan to investigate how the model applies to the persons with impaired balance. The investigation may show how the measurements resulting from the model correlate with performance of balance. Our model indicates a time delay associated with standing. Time delays may be an indicator of system performance and stability. Exploration of the effect of time delay is also planned as a topic for future studies.
[047] THE EFFECT OF PELVIC OBLIQUITY AND STANCE-PHASE KNEE FLEXION ON THE VERTICAL DISPLACEMENT OF THE TRUNK DURING NORMAL WALKING
Dudley S. Childress, PhD; Steven A. Gard, PhD
Northwestern University Rehabilitation Engineering
Research Program, Chicago, IL 60611; email:
d-childress@nwu.edu;
Web: http://www.repoc.nwu.edu/
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 22202
PURPOSE--Conventional wisdom in orthopedics states that the function of pelvic obliquity and stance-phase knee flexion in normal walking is to lower the trunk at its peak vertical displacement, thereby reducing the trunk's vertical excursion and minimizing the energy expenditure of gait. However, kinematic gait data from normal ambulators indicate that these movements are maximum at about the time of toe off and are nearly neutral by the time the trunk reaches its peak vertical displacement during the gait cycle. Therefore, we believe that the roles of pelvic obliquity and stance-phase knee flexion in normal walking may be different from what is generally believed. We measured pelvic obliquity and stance-phase knee flexion in normal ambulators, and calculated the effect of these movements on the trunk's vertical displacement.
METHODOLOGY--We acquired data from three adult male controls with no known gait pathologies at the Children's Memorial Hospital Gait Analysis Laboratory in Chicago, a laboratory that routinely performs clinical gait analyses. Kinematic measurements were made with a Vicon system, and ground reaction forces were measured with two AMTI forceplates embedded in the walkway. The subjects walked at four speeds from about 1.0-2.4 ms. For each of the subjects, three data trials at each of the walking speeds were acquired and processed for our investigation.
RESULTS--Our results showed that in normal walking pelvic obliquity is maximum at about the time of toe-off, and is nearly neutral by midswing. We determined that pelvic obliquity has negligible effect on the vertical excursion of the trunk, which is contrary to what has been widely reported in the literature. The reason for this discrepancy is because pelvic obliquity is out of phase with the trunk's vertical displacement. We observed that the primary effect of pelvic obliquity on the trunk's vertical displacement was to decrease its mean elevation by about 2-4 mm and shift the waveform slightly in phase from where it would have been otherwise. Because normal pelvic obliquity is initiated with heel contact during gait, and reaches a maximum near contralateral toe-off, we believe that pelvic obliquity serves as a physiological shock absorber to reduce ground reaction forces as load is shifted from the trailing to the leading leg, a role which has been suggested in the literature by other researchers. Stance-phase knee flexion in normal walking was found to decrease the vertical position of the trunk by about 10 mm at the slowest walking speeds, and up to 45 mm at the highest speeds. However, as with pelvic obliquity, we observed that the vertical displacement due to stance-phase knee flexion was out of phase with the trunk's vertical displacement due to limb geometry during gait. As a result, the effect of stance-phase knee flexion on the vertical excursion of the trunk was found to be minimal. The phase relationship that stance-phase knee flexion has with the trunk's vertical displacement waveform is nearly identical to that we observed for pelvic obliquity. Based upon our results we believe that, like pelvic obliquity, stance-phase knee flexion plays a critical role in shock absorption during normal walking.
FUTURE PLANS--Both stance-phase knee flexion and pelvic obliquity appear to play significant roles in reducing the magnitude of the force transmitted from the ground to the body in normal walking. By designing prostheses and prosthetic components that will restore, or simulate, the shock-absorbing functions of these physiological movements, we believe that the gaits of persons with amputation may be substantially improved. Implementing new prosthetic designs and fitting techniques based upon our results will involve educating rehabilitation clinicians on the shock-absorbing nature of pelvic obliquity and stance-phase knee flexion, since much of the literature appears to have misinterpreted their role in normal walking.
[048] GAIT ASSESSMENT FROM SINGLE MARKER MEASUREMENTS
Dudley S. Childress, PhD; Richmond B. Chan, PhD
Northwestern University Rehabilitation Engineering
Research Program, Chicago, IL 60611; email:
d-childress@nwu.edu;
Web: http://www.repoc.nwu.edu/
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 22202
PURPOSE--Velocity, cadence, and step length have always been important measurements associated with human walking and with assessment of human gait. In this project, we investigated the efficacy and applicability of a simple method of gait analysis that produces several parameters used in gait analysis through the analysis of one marker placed near the body center-of-mass (COM). This method is not a replacement for conventional gait analysis, but may have clinical applications. This work forms the instrumentation foundation for our development of simple methods of gait assessment, such as the Direct Ultrasound Ranging System (DURS).
METHODOLOGY--A marker was placed on the subject's trunk at near the S-2 level. Data were collected and analyzed as several subjects walked naturally. Although simpler instrumentation system may be used, the CODA-3 movement monitoring system was used in the study because of its availability.
PROGRESS--A single marker placed near the COM at the level of S-2 can provide information on the velocity, step length, cadence, spatial and temporal symmetry, rate to achieve steady state walking, lateral and vertical body movement, and some gait initiation parameters. Velocity data were obtained by taking the derivative of position data. Analysis of the occurrence of peaks and valleys of the waveforms from vertical position and forward velocity of the COM provided information on cadence, step length, step duration, and the number of steps to achieve steady walking speed. Temporal and spatial symmetry was obtained by comparing waveforms of successive steps. Gait initiation is a topic of interest by several investigators. Some important parameters related to gait initiation that may be obtained using one-marker analysis include the time and distance needed to achieve steady state walking speed. Passing the instantaneous velocity and acceleration profiles through a very low pass filter provides the profiles of the mean walking velocity and acceleration. These profiles provide information about the kinematics, time and distance necessary to achieve steady-state walking speed from standing.
RESULTS--This study shows the theoretical rationale for one-marker analysis, as well as its convenience, practicality, and usefulness. The amount of data that needs to be gathered in this kind of analysis is minimal, allowing speedy data processing and analysis. We do not suggest that one-marker analysis provides all information on walking, or should replace other existing means of gait analysis, but that it should be seen a convenient means of obtaining practical and useful information about gait. Several results of this work have been applied to DURS.
FUTURE PLANS--Although this method is simple and provides a lot of useful information, further studies to confirm the reliability of the method are needed. We plan to develop models of gait and relate the measured data to the models. Investigation of the clinical significance of the waveforms derived from one-marker analysis is also planned. We will use the results of this work to enhance the development of the DURS.
[049] USE OF JOINT TORQUE, ENERGY, AND POWER IN CLINICAL GAIT EVALUATION
Sheldon R. Simon, MD; Necip Berme, PhD; Mark Geil
The Division of Orthopaedics, and the Department of
Mechanical Engineering, The Ohio State University, Columbus,
OH 43210; email: Simon.1@osu.edu; Berme.1@osu.edu;
Geil.1@osu.edu
Sponsor: National Institute for Disability and Rehabilitation Research, Rehabilitation Engineering Research, Washington, DC 22202
PURPOSE--Clinical evaluation of the effectiveness of prosthetic feet is essential as new technologies emerge. Recent development in lower-limb prosthetics has been aimed at feet that store energy in mid-stance and return the energy at toe-off to assist the user in propulsion. Joint power is used to determine the amount of energy stored and returned, which is then used to determine a specific foot's clinical efficacy. This project will address the limitations in current methods for power calculation and explore improved methods.
METHODOLOGY--Inverse dynamics has been the primary analysis model used in gait analysis to estimate the net mechanical power and work required to move the body through a gait cycle. However, researchers have recently pointed out the incorrect assumptions of inverse dynamics when applied to ankle-foot devices: the model assumes a rigid foot articulating about the ankle, and there is no accounting for the energy dissipation and recovery in the viscoelastic material of the cosmetic cover.
The Dynamic Elastic Response (DER) foot used for the study contains three carbon-fiber deflection plates for energy storage. Materials testing was performed to determine the mode of deformation and the damping characteristics of the foot. In order to determine the presence of hysteresis or viscoelasticity, the foot was loaded cyclically between 10 and 800 N at 0.5 Hz. A portable x-ray unit was positioned in front of the MTS machine to record medial-lateral images of the foot at various stages of deformation. A known reference length bar was positioned in the field of view to normalize varying degrees of magnification. Images were recorded at these approximate loads: unloaded, 200 N, 300 N, 400 N, 685 N (approximating body weight), and 800 N. The films were digitized and processed using the public domain NIH Image program to obtain deflection angles at incremented vertices along the length of each deflection plate.
PROGRESS--The load-deflection curve for the foot under cyclic loading between 10 and 800 N shows that once a consistent cycle is reached, the loading-unloading cycles form a loop, showing the presence of hysteresis in the foot structure. Representative energy loss in the hysteresis loop for a single cycle was 1,128.31 N-mm. This figure is 14.9 percent of the input energy.
When compared to the image for the unloaded foot, measurement of angle at the vertices superimposed on the large plate shows that deflection occurs along the length of the plates as in cantilever beam loading. Similar results were obtained for each plate at each loading increment. Furthermore, at maximum deflection, the ankle joint center of rotation assumed by inverse dynamics is actually the only rigid portion of this DER foot.
RESULTS--Hysteresis in the foot structure confirms the presence of energy dissipation with each cycle, most likely due to the cosmetic foam cover. Energy dissipation is not accounted for in the inverse dynamics approach. Its presence in the gait cycle can result in erroneous conclusions regarding energy stored and returned by the prosthetic foot. As this factor is prevalent in the literature as a means for comparison of different DER feet, this represents a serious inadequacy.
[050] DEVELOPMENT OF A GAIT INTERPRETATION, INSTRUCTION AND REPORT GENERATION SYSTEM
Sheldon R. Simon, MD; Philip J. Smith, PhD; Rebecca J.
Denning, PhD
The Division of Orthopaedics, and the Cognitive Systems
Engineering Laboratory, The Ohio State University, Columbus,
OH 43210; email: psmith@magnus.acs.ohio-state.edu;
rdenning@magnus.acs.ohio-state.edu
Sponsor: National Institute for Disability and Rehabilitation Research, Rehabilitation Engineering Research, Washington, DC 22202
PURPOSE--Both health care and rehabilitation for orthopedic clients may be improved by taking advantage of the increasingly sophisticated quantitative measures now available for gait analysis. To ensure the effective use of these measures, our goal is to develop a computer-based tutoring and report-generation system. This system can be used to help orthopedic residents and physical therapists gain experience and skill in analyzing various gait dysfunctions, and to develop more informative patient reports for referring physicians.
METHODOLOGY--The Dr. Gait III tutor and report generation system is currently being developed on a Macintosh computer environment using the C programming language. This environment allows users to manipulate multimedia elements such as sound, video, animation, illustrations, and graphical representations easily. Providing data in these formats will support both the instruction of individuals interested in acquiring the skills of gait analysis and the production of detailed client reports for those who manage and provide care.
PROGRESS--A fully functional prototype of Dr. Gait III has been developed and is being subjected to beta testing by orthopedic surgeons and residents, physical therapists, and physical therapy students. This system will also be ported to an IBM environment as evaluation results are compiled and the preliminary system design refined.
The system allows users to enter, review, and annotate data from actual cases. This annotated data and accompanying text segments can be incorporated into a detailed report of progress. Preliminary interviews with orthopedic surgeons revealed that in addition to providing a tutoring system for residents, the ability to maintain and manage client information is an important additional function.
Each screen of data may be annotated in a manner appropriate to the type of data. Text screens provide the capability to change the style (bold, italics, etc.) and color of the text. The colors available have been chosen for saturation and depth to be distinctly identifiable when printed on a grayscale printer.
We are currently developing a suite of teaching cases for which the tutoring aspects have already been authored. In addition, the system may be used to author new cases for which no tutoring has been recorded.
Any item of data, along with its annotations, may be selected for inclusion in the report. This provides a simple method for including annotated data that is not very time-consuming, because many of the annotations can be made automatically. Additionally, text can be generated to accompany some of the findings. Once the data has been noted for inclusion in the report, a procedure is run to build the report in an organized manner based on common templates in a word processor. Further editing of the report is still necessary, but the data is already in place, making it easy to view the substance while writing the discussion.
FUTURE PLANS--Following the completion of the current evaluation, a formal evaluation study is planned. Following this evaluation, the Dr. Gait III system will be distributed for general use as both a case-based interactive learning environment and a client evaluation and care management tool. Further evaluations will accompany this distribution.
[051] REFINEMENT, EVALUATION, AND DISSEMINATION OF A DIAGNOSTIC AND TREATMENT ASSESSMENT EXPERT SYSTEM FOR THE INTERPRETATION OF WALKING DISORDERS LEADING TO DISABILITY
Sheldon R. Simon, MD; Jack W. Smith, MD; Kathy A.
Johnson, PhD
The Divisions of Orthopaedics and Medical Informatics
The Ohio State University, Columbus, OH 43210; email:
Simon.1@osu.edu; Smith.30@osu.edu; Johnson.32@osu.edu
Sponsor: National Institute for Disability and Rehabilitation Research, Rehabilitation Engineering Research, Washington, DC 22202
PURPOSE--This project involves the continuing development of a diagnostic and treatment expert system for the interpretation of walking disorders leading to disability. Over the past decade, research efforts have been developing tools to assess human gait performance objectively. To assist clinicians, computational methods in analyzing gait could ensure a standardized, high quality level of analysis, decrease the time involved in doing an analysis, provide report-writing support, provide updates for new techniques, and be part of a tool for instruction on gait analysis.
METHODOLOGY--We have been working with a prototype expert system called QUAWDS (Qualitative Analysis of Walking DisorderS) for diagnosing cerebral palsy gait disorders from the multiple sources of raw data used by the gait analysis expert. The major subtasks are: motion deviation identification, muscle fault generation, muscle fault rating, explanatory coverage determination, and determination of overall interpretation. These modules use a combination of associational knowledge (rule-like) and a qualitative model of the physical system.
We are in the process of isolating these and embedding them as decision-support tools within a user-friendly cooperative problem-solving interface, so that a gait analysis expert can easily use any or all submodules of QUAWDS for gait analysis. We are also evaluating the system's performance against human experts, refining the decision-support tools to achieve expert level performance.
PROGRESS--We have nearly completed the transfer of knowledge from QUAWDS to decision-support tools to be embedded in the system. Currently, there are tools for identifying significant findings with respect to joint angle graphs, range of motion, time and distance data, and EMGs. We have programmed additional functions that can be used to pull information from the AI decision-support tools into a report. For example, the motion findings can be output in both textual and graph form, and the text has been refined to read in a more human-like manner so that they can be incorporated into a report with little or no modification. We have also written code to determine the muscle faults associated with a particular deviation and we are currently developing the interface to integrate this function into the interface. Some of the errors in QUAWDS' performance were known to be related to the method of motion finding determination; therefore, we have been experimenting with alternative methods of finding determination such as Sutherland's prediction regions.
RESULTS--The system is being beta-tested at several sites. Providing a single place for data to reside for gait analysis is proving to be quite a useful concept for both teaching gait analysis and facilitating clinical report generation. The decision-aid tools that have been provided up to this point seem to be useful and fairly accurate. Another benefit is the ability to explore other methods for accomplishing subtasks of gait analysis and comparing the outputs easily. This eventually leads to a better method for performing gait analysis as well as better support for the people currently doing (or learning to do) that analysis.
FUTURE PLANS--We are working to add new functionality to the system based on other research into the clinical significance of data analysis types for gait analysis. As we determine new categories of findings based on our research into torques, powers, moments, time/distance parameters, and the like, decision-aid tools will be developed and added to the system. These tools will improve both the functioning of the system as well as our understanding of the process of gait analysis.
[052] THE INFLUENCE OF FATIGUE ON EMG AND IMPACT ACCELERATION IN RUNNING
Joseph Mizrahi, DSc; Arkady Voloshin, PhD; Dalia Russek,
BSc; Oleg Verbitsky; PhD; Eli Isakov, MD
Department of Biomedical Engineering, Technion-Israel
Institute of Technology, Haifa, 32000 Israel; Department of
Mechanical Engineering and Mechanics, Lehigh University,
Bethlehem, PA, 18015, USA; Loewenstein Rehabilitation
Center, Raanana 43100, Israel; email:
JM@biomed.technion.ac.il
Sponsor: The Segal Foundation, Israel
PURPOSE--This research was aimed at analyzing the effects of fatigue on the ability of the human musculoskeletal system to attenuate the heelstrike-generated shock waves in running and to correlate any existing changes in the EMG signal with other fatigue measures.
METHODOLOGY--Each subject is instrumented with a light-weight accelerometer placed externally over the tibial tuberosity and with two pairs of bipolar surface EMG electrodes placed over the quadriceps and gastrocnemius. The subjects run on a treadmill for 30 min at a speed corresponding to their individual anaerobic threshold level, while the acceleration and EMG data are sampled. Fatigue is monitored by the respiratory end-tidal CO;i2 pressure.
Data processing is simplified, due to the constant speed of running during each test. Software is developed that utilizes the given speed and location of the first heel strike. After this, it automatically processes and detects all consequent occurrences of heel strike.
The analyzed variables are the maximum amplitudes of the accelerations recorded at foot strike on the tibial tuberosity; mean and RMS of the EMG signal in the time domain, mean and median frequency of the same signal in the frequency domain, and PETCO;i2. For each subject, the maximum accelerations recorded in one file are measured and averaged. After this, the acceleration data are normalized with respect to the record at the beginning of the experiment.
The obtained data are processed via conventional statistical routines available in "MATLAB." Differences between results were tested (t-test) and the level of significance determined at p<0.05.
RESULTS--The obtained results show that whenever general fatigue occurred, the amplitude of acceleration signal steadily increased. However, in both the time and frequency domains, EMG did not correlate with fatigue. Thus, one may conclude that when fatigued, the human musculoskeletal system becomes less capable of handling the heelstrike-induced shock acceleration.
FUTURE PLANS--It is intended to extend this study to include different fatiguing protocols. EMG measurements will be extended to include the tibialis anterior and hamstrings muscles, in addition to the already monitored gastrocnemius and quadriceps muscles. Kinematic measurements will also be made during the experiments. A biomechanical model of the lower limb will be developed by which the variations of the elastic and damping coefficients of the segments can be evaluated as a function of the fatigue level.
[053] ASSESSMENT OF AMBULATION MOTION PARAMETERS FOR CLINICAL EVALUATION
Sheldon R. Simon, MD; Necip Berme, PhD; Tasos
Karakostas
The Division of Orthopaedics, and the Department of
Mechanical Engineering, The Ohio State University, Columbus,
OH 43210; email: Simon.1@osu.edu; Berme.1@osu.edu;
Karakostas.1@osu.edu
Sponsor: National Institute for Disability and Rehabilitation Research, Rehabilitation Engineering Research, Washington, DC 22202
PURPOSE--The overall goals of this work are to develop simple motion parameters and indices for general and specific patient populations; evaluate the computational complexity of such parameters while investigating sources of error and variability; examine the feasibility of implementing these parameters in clinical data processing; assess the usefulness of these parameters in diagnosing motion disorders, prescribing interventions, evaluation of long term prognoses, and assessing disability levels; and finally, develop a database of normal values for these parameters.
METHODOLOGY--At this phase of the project, we are developing parameters that will assist orthopedic surgeons who perform musculotendon lengthening in children with cerebral palsy (CP) in determining the appropriate amount of lengthening. Following appropriate modeling of the kinematics of each of the joints of the lower limb, the kinematic and kinetic information with respect to the musculotendon units of interest (e.g., moments, moment arms, forces and lengths, and strains) can be determined at different joint positions. The kinetic data can be estimated as they relate to one or several musculotendon units. Our analysis indicates that the length and force outputs determined for each individual musculotendon unit are the most acceptable variables to be investigated. We have also been able to use our already existing database as well as information collected from recent studies in an effort to combine pre- and postsurgery data to investigate the effects of musculotendon unit lengthening.
PROGRESS--We are incorporating motion data from gait evaluations from persons with CP who are to undergo different treatment procedures, such as either the use of ankle foot orthotics or derotational osteotomies, to determine the effects of these treatments also. Our initial results indicate that surgery, and in particular musculotendon lengthening, has the most dramatic effects.
In addition, in order to extend the capabilities and theoretical reliability of the musculoskeletal model, we have proceeded in specific modeling of the knee joint. Our approach involves consideration of equilibrium conditions when solving for the knee joint moments, moment arms, and musculotenon forces. The mathematical formulation of the improved knee joint model has been completed.
In order to establish a database for the muscle strengths of individuals without any pathology for comparison, we performed gait evaluations on 14 controls under different conditions, rested and fatigued. Our results indicate that individual musculotendon groups of interest operate at specific regions on Hill's curve.
Another objective of this project has been the determination of the threshold level of spasticity. We have established the importance of the careful processing of the electromyographic signal, demonstrating that even with maximum isometric contractions, the signal will vary depending on the task that the specific musculotendon unit is required to perform.
However, accurate identification of the threshold level of spasticity, and accurate identification of the individual muscle strengths presupposes the ability to estimate individual muscle forces. Toward this end we have constructed an EMG-driven model whose final output is a cocontraction, or spasticity, index. Our theoretical and mathematical formulation of the EMG-driven model has been completed. and it is being implemented in computer code. In the very near future data will be collected from maximum of fourteen subjects to validate the model.
[054] EFFECT OF AN INDUCED LEG-LENGTH DISCREPANCY ON GAIT BIOMECHANICS
Alberto Esquenazi, MD; Mike Hatzakis, MD; Mukul Talaty,
MS, BME
Gait amd Motion Analysis Laboratory, MossRehab Hospital,
Philadelphia, PA 19141; Thomas Jefferson University
Hospital, Philadelphia, PA 19107;
email: aesquena@aehn2.einstein.edu
Sponsor: None listed
PURPOSE--A major goal is to better understand the physiologic factors involved in the compensation for the inequality of lower limb lengths. It is desired to document at what levels of leg-length inequality does compensation at the hip occur, and what overall changes occur in the mechanics of gait. With this information, pathologic changes that occur in the hips, lower back, and lower limbs due to compensation for leg-length inequality may be better understood. This may facilitate inferences on effective treatment options.
METHODOLOGY--Six to 10 controls with documented normal leg length (using both the ASIS to floor method and ASIS to medial malleolus methods to assure no structural or functional pre-existing pathology) will be tested by completion of the study. Many have studied the clinical assessment of degree of leg-length discrepancy, and we choose the ASIS to medial malleolus because it seems to be the most rigorously tested. While some claim that ASIS to lateral malleolus can detect functional leg-length inequalities, this technique has not been as well demonstrated in the literature. Two mechanisms of leg-length discrepancy are used: a tapered heel lift (variable thickness with maximum thickness of 1.27 cm at the heel) is placed in one shoe, and a full shoe lift (constant thickness of 1.27 cm) is placed under the shoe. Each subject is given a chance to accommodate to the test condition before data are recorded and is tested without any orthotic as well. Temporo-spatial foot fall parameters (such as stride lengths and times, stance/swing lengths and times, step lengths, and walking velocity) as well as ipsilateral hip, knee and ankle kinematics, and kinetics are recorded. Upon completion of the test session, the subject is given the partial lift to wear during all activities for 2 to 3 wks. Each subject is then re-tested as outlined above in just the heel lift and no lift conditions.
PROGRESS--Several subjects have been tested to date; testing and analysis is ongoing.
PRELIMINARY RESULTS--There appear to be statistically significant differences in lower limb kinematics and kinetics of the longer limb. Differences are noted between the lifted limb and same limb when no orthotic is used to induce a leg-length discrepancy, as well as between the lifted limb before and after the accommodation period.
Specific changes for a very limited number of subjects have been noted in pelvic obliquity and range of motion, pelvic rotation, hip flexion, hip adduction and range of motion, hip rotation, and various forces. Some of the changes seem to correlate with the expected responses to a lift (such increased ipsilateral hip flexion and pelvic obliquity) where as others do not. There is undoubtedly a complex series of interactions that give rise to the mechanisms individuals use to compensate for the intervention. An insufficient number of subjects has been tested so far to be able to speculate about trends in the methods of compensation.
FUTURE PLANS--Differences between short- and long-term changes in the biomechanics of gait in compensation of leg-length discrepancy and the characteristics of leg-length discrepancy that lead to pathologic changes in the hip and back are areas of focus.
[055] WHEELCHAIR PROPULSION PERFORMANCE IN YOUNG, MIDDLE-AGED, AND ELDERLY
Mary M. Rodgers, PT, PhD; Peter Gorman, MD; Randall
Keyser, PhD; Pamela Russell, PhD
VA Medical Center, Baltimore, MD 21201; email:
mrodgers@physio.ab.umd.edu
Sponsor: Department of Veterans Affairs, VA
Rehabilitation Research and Development Service, Washington,
DC 20420
(Project #B764-2RA)
PURPOSE--The purpose of this program is twofold: 1) to investigate how age affects wheelchair performance for lower-limb impaired wheelchair users; and 2) to test a specific exercise intervention for its effectiveness in reducing potentially injury-producing biomechanical characteristics and excessive physiologic stresses.
METHODOLOGY--Anthropometric and muscle strength measurements, neuromuscular assessments, and wheelchair propulsion testing were performed before and following exercise training (stretching, strengthening, and aerobic training three times weekly for 6 wks). Shoulder, elbow, and wrist joint kinetics (joint moments and joint reaction forces) were calculated from 3-D motion and handrim force data. Changes with training in all variables were statistically tested using repeated measures ANOVAs (pre/post vs. unfatigued/fatigued) with a 0.05 level for significance. Interaction effects were further analyzed with paired t-tests and a Bonferroni correction for multiple comparisons.
PROGRESS--Twenty wheelchair users completed the study. Eleven additional subjects completed pretesting only. Seven nonusers (controls) have been tested for database comparisons.
PRELIMINARY RESULTS--Twenty subjects (age 44±11 yrs; weight 78±20 kg; spinal cord lesion level T2-L5 and multitrauma; 16 male, 4 female; wheelchair users for 16±10 yrs) completed the study. Significant results of the training intervention on all measures follows. Anthropometric and temporal measures were unaffected by training. Eccentric isokinetic strength increased for wrist flexors (38.5 percent), extensors (27.8 percent), radial deviators (30.1 percent) and ulnar deviators (32.5 percent). Joint kinematic changes included a 3.2 percent increase in maximum elbow extension and a 6.5 percent increase in maximum trunk flexion during the entire wheelchair propulsion cycle. The increase in trunk forward flexion was evident during both contact (6.1 percent) and time of release (6.9 percent), and increased 8.9 percent in the fatigued condition at the post-test compared to the pretest. Handrim kinetic changes included an increase in the maximum propulsion moment (1.1 percent) and in the resultant moment (1.2 percent). Joint kinetic changes included an increase in the shoulder flexion moment (Mz) of 16.2 percent and shoulder anterioposterior shear force of 16.4 percent. The timing of muscle activity as determined from EMG changed significantly for the middle trapezius which came on 16.1 percent later in the cycle and reached peak activity 13.1 percent later, and shut off 10.9 percent later.
FUTURE PLANS/IMPLICATIONS--Initial findings indicate that specific training for wheelchair users can improve wheelchair propulsion mechanics, possibly decreasing susceptibility to overuse injury. Further investigation of joint stresses is continuing. Three additional controls will be tested before completion of the study for comparison. A videotape of the exercise training program will be produced.
[056] QUANTIFICATION OF MUSCULAR AND TENDINOUS FORCES OF THE HAND DURING STATIC EFFORT
Daniel Bourbonnais, PhD; Sylvain Gagnon, Dr; André
Chollet, Dr; Pierre Duval, Dr, Eng; Martin Le Blanc, Eng
Centre de Recherche de l'institut de réadaptation de
Montréal, Montréal, QC Canada, H3S-2J4
Sponsor: Conseil de recherches médicales du Canada; Wright Medical Technology
PURPOSE--The purpose of this project is to develop a valid biomechanical model of the human finger, quantifying the muscular and tendinous forces and joint constraint forces exerted on the structures of the hand during static efforts.
METHODOLOGY--We propose a new method to quantify muscle forces based on in vivo measurements and an anatomical model of the finger. The experimentation will be realized in persons receiving a bone plate following a transverse fracture of the metacarpal or phalangeal bone. The regular bone plate will be temporarily removed and replaced by a sensor, allowing measurements of force and moments of force exerted at the site of the fracture in four degrees of freedom (forces in tension-compression and moments of force in the ventro-dorsal, medio-lateral, and rostro-caudal axes of the bone). The subjects will receive a block of the nerves innervating the tested finger and will be asked to perform static efforts at the distal tip of the finger in the flexion-extension plane, while forces and moments of force on the bone will be measured. These measurements will be obtained with the finger in two positions, that is, the metacarpo-phalageal joint (MCP) in extension and in flexion. A lateral and dorsal X-ray of the finger will be realized to quantify the location of the transducer relative to bone structures. Once the measurements are completed, the sensor will be removed and the surgery completed as usual. Using the force and moment of force values measured during the surgery and moment arm values estimated using the geometrical model of the finger, it will be possible to solve independent equations of static equilibrium and to quantify muscular and tendinous forces.
PROGRESS--To date, the experiments have been conducted on 10 persons, 4 in the preliminary experiments and 6 using the new protocol and apparatus. The force in compression, as well as the moments of force were measured in three axes, since a gap between bone segments ensures that these forces and moments of force are transmitted through the transducer. In order to compare the results across subjects, the forces and moments of force measured by the transducer were expressed as in units of applied force at the fingertip. We are now analyzing the results and calculating muscle forces based on the geometrical model of the finger.
PRELIMINARY RESULTS--The results in all subjects indicate that an increasing compressive force and flexor torque were developed progressively as the external force exerted at the tip of the finger increased. The ratio of compressive force per unit of distal force measured at the phalanx and metacarpal varies between 3.7 and 9.8 with a mean of 5.4 (SD=2.2) during a isometric flexion with all joint in extension.
Using these in vivo measurement and our anatomical model of the finger, the values of the muscular and tendinous force acting on the bone are slightly higher than those estimated from other studies using biomechanical modeling of the finger an optimization technique.
FUTURE PLANS--It is expected that this model will provide major clinical impacts in diverse areas, such as hand surgery and rehabilitation. As short-term goal, we believe that repeating the experiments in six to eight additional subjects would help to precisely establish the torque and force patterns. As a long-term goal, it is planned to evaluate or optimize the design of bones, plates, or hand implants using this model.
[057] DETERMINING SOFT TISSUE MATERIAL PROPERTIES FOR THE PURPOSE OF FINITE ELEMENT MODELING OF THE RESIDUAL LIMB IN TRANSTIBIAL AMPUTATION
Dudley S. Childress, PhD; Jeffrey J. Berkley
Northwestern University Rehabilitation Engineering
Research Program, Chicago, IL 60611; email:
d-childress@nwu.edu;
Web: http://www.repoc.nwu.edu/
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 22202
PURPOSE--Finite element (FE) modeling is a potential tool for designing improved prosthetic sockets, based on a prescribed socket/residual limb stress distribution. Currently, the accuracy of modeling the socket/residual limb system has been limited by inaccurate representation of soft tissue material properties. An indentor test has been used by previous investigators to estimate bulk tissue modulus values at various locations, but the accuracy of this method is limited because only a small number of sites can be tested. Also, the indentor test only gives an average bulk modulus at the test site and yields no information on how material properties vary radially. Our goal is to demonstrate that we can use CT scans and their grayscale values to obtain a good estimate of tissue modulus values throughout the whole limb.
METHODOLOGY--It is our hypothesis that a relationship can be found between CT scan grayscale values and the Young's modulus of the soft tissue. The grayscale distribution method involves using force/displacement data obtained from CT scans of a residual limb deformed by an ICECAST pressurized casting apparatus. Using metallic surface markers, CT scans can be used to track hundreds of surface points. An FE model with surface nodes that correspond to marker locations can be developed from the CT scan images. Testing average gray scale value is found for each element and an estimate of Young's modulus is assigned. The experimental loading conditions are then simulated and the FE surface marker displacements are compared to the experimental surface marker displacements. The coefficients of the transformation equation are optimized, based on the surface marker displacement error. The purpose of this study is to find the transformation relationship that will most accurately convert grayscale values to Young's modulus values.
PROGRESS--Thus far, theoretical modeling and testing of a physical model have been completed. Theoretical modeling of a residual limb was used to see whether optimization could be used to find material properties using a surface displacement error criterion. With theoretical modeling, no experimental measurements were obtained. Instead, material properties are distributed through a simulated experimental FE model and "experimental" displacements are calculated. Using these displacements, an attempt was made to solve the material properties of the model. The success of the solving procedure was then evaluated.
A physical model was developed to further test our procedure for solving material properties. This type of testing differed from the theoretical model testing in that external sources of error were present, including digitizing error, inaccurate boundary condition representation, and error inherent to FE modeling. The objective was to see whether the material-property-solving procedure would be adversely affected by these error sources. The physical model was based on homogeneous materials with known material properties, to allow comparison with the material property solutions. CT scans taken of the model under various pressures and displacements were measured. The optimization procedure was applied and the accuracy of the solution assessed.
RESULTS--Theoretical model testing proved the potential for using an optimization procedure to solve material properties based on surface displacement error. Also, the physical model testing proved successful. The material property of the physical model was solved within 1.5 percent of the actual value. The overall surface displacement error was less than 17 percent, when errors were weighted by the magnitude of their displacements. When all surface errors are weighted equally, the surface displacement error increases to 44 percent. Areas that had small displacements showed higher errors, indicating that digitizing error can adversely affect results. If the digitizing error is of the same magnitude as the displacement, predictive accuracy deteriorates.
FUTURE PLANS--Our next test will be conducted on a human subject. We have received Institutional Review Board approval and plan to initiate testing in the near future.
[058] STANDING SWAY: KINEMATICS AND DYNAMICS FROM FORCEPLATE MEASUREMENTS
Joseph Mizrahi, DSc; Oron Levin, MSc; Moshe Shoham
Departments of Biomedical and Mechanical Engineering,
Technion-Israel Institute of Technology, Haifa, 32000
Israel; email: JM@biomed.technion.ac.il
Sponsor: The Segal Foundation and the Fund for the Promotion of Research at the Technion, Israel
PURPOSE--We are working to develop a model for the estimation of the kinematics and dynamics of the lower extremities in standing sway from bilateral forceplate measurements.
METHODOLOGY--A three dimensional, five-segment four-joint model of the human body is used to describe postural standing sway dynamics. Force-plate data of the reactive forces and centers of pressure are measured bilaterally. By applying the equations of motion to these data, the transversal trajectory of the center of gravity (CG) of the body is resolved in the sagittal and coronal planes. An inverse kinematics algorithm is used to evaluate the kinematics of the body segments. The dynamics of the segments are then resolved by using the Newton-Euler equations, and the model's estimated dynamic quantities of the distal segments are compared with those actually measured. Differences between model and measured dynamics are calculated and minimized using an iterative algorithm to re-estimate joint positioning and anthropometric properties.
RESULTS--The above method was illustrated on a group of 11 nonimpaired controls. The average relative errors were: 0.34 (±0.25) percent for the vertical forces; 1.97 (±1.12) percent for the ML forces and 15.7 (±14.8) percent for the torques in the sagittal plane. Relatively large errors were found for the AP forces 62.7 (±64.2) percent and the torque in the coronal plane 33.4 (±43.8) percent. The angular displacements in the right ankle were found to be in opposite phase with the right hip, both in the coronal and sagittal planes. Similar angular displacements were found on the left side.
The results indicate that the relative errors obtained in the final iteration are of the same order of magnitude as those reported for closed loop problems involved in direct kinematics measurements of human gait.
FUTURE PLANS--It is intended to extend the developed model to enable evaluation of the muscle forces crossing the joints. EMG measurements of the major muscles of the lower limbs will be performed and incorporated in the solution procedure of the model.
[059] BILATERAL DEFICIT DURING KNEE EXTENSION
Brian Woods; J.R. Sexsmith, PhD; D.F. Lovely, PhD
Institute of Biomedical Engineering, University of New
Brunswick, Fredericton, NB CANADA E3B 5A3; email:
biomed@unb.ca
Sponsor: None listed
PURPOSE--The project studied the reproducibility, reliability, effect of contraction velocity, and neuromuscular control of bilateral and unilateral maximal voluntary contractions.
METHODOLOGY--In order to understand the coordinated functioning of the neuromuscular system, we need to know how it functions during a multitude of specific conditions, such as when a coordinated task involves one limb (unilateral) as compared to the same task performed with both limbs simultaneously (bilateral). To quantify the relative contributions of the unilateral and bilateral performances bilateral ratios (BLR) are calculated.
RESULTS--The reproducibility, reliability and the effect of contraction velocity on BLRs during dynamic omnikinetic maximal knee extensions were assessed. In summary, the peak torque BLRs and knee angle BLRs were similar to previous literature which employed isometric contractions. In addition the intra- and inter-test reliability of all BLRs was extremely high (r>0.95). Contraction velocity caused BLR to decrease; that is, the force generated during bilateral contractions decreased more than during unilateral contractions. Furthermore the relationship between knee angle and BLR was perturbed by contraction.
FUTURE PLANS--We plan to investigate the neuromuscular control differences between bilateral and unilateral contractions by recording myoelectric signals from the quadriceps and hamstring muscles.
Go to top.
|
Previous |
Contents |
Next |
|---|
Last revised Fri 04/30/1999