IEEE Transactions on Rehabilitation Engineering

Published by Institute of Electrical and Electronics Engineers
Online ISSN: 1063-6528
Publications
Article
Interface pressures and shear stresses were measured at 13 sites on two unilateral below-knee amputee subjects ambulating with lower-limb patellar-tendon-bearing prostheses. Interface stresses at the time of the first peak in the shank axial force-time curve were investigated at different socket-shank alignment settings. Stress magnitudes ranged from 1.2 to 214.7 kPa for pressure and 0.4 to 79.6 kPa for resultant shear stress, and changes in stress due to misalignment ranged from 1.3 to 80.7 kPa for pressure and from 0.2 to 38.0 kPa for resultant shear stress. For both subjects interface stress changes were much greater in the anterior socket region than in the lateral or posterior regions. Thus, alignment changes had a localized effect on interface stresses. Plots of alignment versus pressure or resultant shear stress were nonlinear for both subjects, in a number of cases maximizing or minimizing at the nominal alignment, indicating complex interface stress-alignment relationships. Variation (standard deviation/mean) was not significantly different for nominal versus misaligned steps, indicating that the subjects adapted well to the alignment changes. Session to session differences in interface stresses were typically larger than interface stress differences induced by alignment modifications. Thus, while these subjects compensated well for alignment changes to maintain consistent interface stresses within a session, they did not do so for different sessions conducted weeks apart.
 
Article
Nocturnal incontinence (enuresis) affects 20% of children over four years old, and this figure typically decreases by 15% each year. At the age of 18, 1% of those people remain enuretic. Nocturnal enuresis can be treated by means of a conditioning device that awakens the patient once the urine level has reached a preestablished threshold of the capacity of his or her bladder. We have designed and implemented a portable miniaturized ultrasonic monitor, which permits estimation of the urine volume with an accuracy of 75%. Prototypes have been completed and validated on 41 patients (children) at Ste. Justine Hospital (Montréal). On the first group of 33 patients, we used a hand-held transducer to determine the volume detection range, which is accurate for volumes between 40 and 400 mL. With the second group of 8 patients, the device was mounted on an elastic belt around the abdomen. Measurements were taken in order to validate the accuracy of urine threshold detection and the activation of the corresponding alarm.
 
Article
Electrotactile (electrocutaneous) stimulation at currents greater than sensation threshold causes sensory adaptation, which temporarily raises the sensation threshold and reduces the perceived magnitude of stimulation. After 15 min of moderately intense exposure to a conditioning stimulus (10 s on, 10 s off), the sensation threshold elevation for seven observers was 60-270%, depending on the current, frequency, and number of pulses in the burst structure of the conditioning stimulus. Increases in any of these parameters increased the sensation threshold elevation. Adaptation and recovery were each complete in approximately 15 min.
 
Article
Functional electrical stimulation (FES) of abdominal muscles as a method of enhancing ventilation was explored in six neurologically intact subjects and five subjects with spinal cord injury (SCI) who had levels of injury between C4 and C7. Pulmonary ventilation was augmented in both groups predominantly due to an increase in tidal volume. The average increase in tidal volume during FES for the neurologically intact group was 350 ml, while in the SCI group it was 220 ml. The FES caused active volume decreases in both the lower thorax and upper abdomen, which together appear to be the mechanism behind the increases seen in tidal volume. Therefore, the proposed method might be useful in future clinical practice. The results indicate that FES of abdominal muscles should be more thoroughly explored as a potential technique of ventilatory support in SCI. The results also point to the necessity for further studies of maintaining the condition of the chest wall in the pulmonary rehabilitation of individuals with tetraplegia.
 
Article
As power sources for rehabilitation equipment, electric, hydraulic, and pneumatic actuators have been used. However a more human-sized and higher powered actuator that can reduce the equipment size is desired. A new metal hydride (MH) actuator that uses the reversible reaction between the heat energy and mechanical energy of a hydrogen absorbing alloy has recently attracted much attention. The MH actuator is characterized by its small size, low weight, noiseless operation and a compliance similar to that of the human elbow joint. Therefore, the MH actuator has the characteristic of being light and easy to use and so is suitable for use in medical and rehabilitation applications. Some lifting devices using this actuator have already been developed and are being used for the care of the aged and disabled. The characteristics of the MH actuator are presented and then some applications are introduced in this paper. It is our opinion that in our aging society the MH actuator will play an important role in the development of medical and rehabilitation equipment.
 
Article
Treating spastic paretic stiff-legged gait, defined as reduced knee flexion in swing, holds a high priority in the rehabilitation of patients with upper motor neuron lesions. We propose a method to determine the relative contributions of hip, knee, and ankle impairments to this disability. We analyzed the gait of ten patients with stiff-legged gait (SLG) due to a single stroke and ten healthy, able-bodied controls. Using subject specific models, we analyzed the induced accelerations (IA's) at the knee. Knee IA's throughout the gait cycle were calculated and the sum of the IA's was compared to the knee joint angular acceleration estimated from kinematic data. The preswing and early swing IA's were the focus of our examination as these largely determine knee kinematics in swing. Knee angular accelerations estimated from IA's and kinematic data agreed for both controls and patients. Gait cycle IA analysis of individual patients identified highly variable causes of SLG including ankle and hip joint impairments. Induced acceleration analysis (IAA) suggested that multiple impairments, not just about the knee, but also about the hip and ankle, lead to this disability. Individual subjects are likely to have individual reasons for their stiff-legged gait. Defining the link between the patients specific impairments and their gait disability should be a goal of clinical gait analysis. IAA is a useful tool for this purpose with a strong potential for clinical application.
 
Article
Detection of knee unlock is a crucial part of finite state artificial reflex control of paraplegic standing supported by functional neuromuscular stimulation (FNS). This paper investigates knee unlock detection schemes using small uniaxial accelerometers mounted on the thigh and shank. Four single and two differential accelerometer configurations were evaluated with respect to knee unlock detection. Their performances were compared to goniometer-based knee unlock detection. Two accelerometers were mounted on the thigh, one just distal to the hip and one just proximal to the knee. The other two accelerometers were mounted on the shank, one just distal to the knee and the other just proximal to the ankle. The sensitive axes of all four accelerometers were directed in the sagittal plane, perpendicular to the long axes of the body segments. The first differential configuration consisted of the two accelerometers on the thigh, the second of the accelerometer on the thigh just proximal to the knee and on the shank just distal to the knee. Detection thresholds were set such that anterior-posterior body sways with locked knees would not yield false knee unlock detections. Experiments were performed in five healthy adult subjects, performing fast and slow knee flexions and anterior-posterior body sways. The first differential accelerometer configuration and the single accelerometer just proximal to the knee gave the earliest detection of fast knee unlocks, earlier than the goniometer-based detection. The second differential accelerometer configuration gave the earliest detection of slow knee unlocks, comparable to the goniometer-based detection. The first differential accelerometer configuration did not provide a reliable detection of slow knee unlocks. The single accelerometer configurations could only detect slow knee unlocks, while rejecting whole body movements, at relatively large knee flexion.
 
Article
Humans can learn to control the amplitude of electroencephalographic (EEG) activity in specific frequency bands over sensorimotor cortex and use it to move a cursor to a target on a computer screen. EEG-based communication could provide a new augmentative communication channel for individuals with motor disabilities. In the present system, each dimension of cursor movement is controlled by a linear equation. While the intercept in the equation is continually updated, it does not perfectly eliminate the impact of spontaneous variations in EEG amplitude. This imperfection reduces the accuracy of cursor movement. We evaluated a response verification (RV) procedure in which each outcome is determined by two opposite trials (e.g., one top-target trial and one bottom-target trial). Success, or failure, on both is required for a definitive outcome. The RV procedure reduces errors due to imperfection in intercept selection. Accuracy for opposite-trial pairs exceeds that predicted from the accuracies of individual trials, and greatly exceeds that for same-trial pairs. The RV procedure should be particularly valuable when the first trial has >2 possible targets, because the second trial need only confirm or deny the outcome of the first, and it should be applicable to nonlinear as well as to linear algorithms.
 
Article
Forces, moments and stresses at the knee are dependent upon external and internal loading factors including muscle forces, segmental position and velocity, load carried, and the moment arms (mechanical advantage) of the muscle-tendon units. Requisite to prediction of forces and moments is a detailed understanding of effective moment arms throughout the knee range-of-motion (ROM). Existing muscle models for the knee are based upon limited static studies of only a few preserved specimens. The objectives of this report are to develop a comprehensive description of muscle-tendon moment arms for the normal knee and the anterior cruciate ligament (ACL)-minus knee during flexion-extension motion. Recent research results describe two nonorthogonal, nonintersecting axes of motion for the knee--one describing flexion-extension (FE) and the other longitudinal rotation (LR, equivalent to internal-external rotation). The effective flexion-extension moment arms of the muscles crossing the knee were developed with respect to the FE axis in 15 fresh, hemi-pelvis cadaver specimens. The normal moment arms for each of 13 muscles plus the patellar tendon exhibited variable, yet repeatable and recognizable patterns throughout the ROM. For most muscles there was no significant difference between the normal and ACL-minus moment arms. The results provide a basis for more accurate predictions of joint reaction forces and moments as well as useful knowledge for practitioners and therapists to assist in the assessment of muscle balance at the knee following injury, repair, and throughout rehabilitation.
 
Article
Individuals with cerebral palsy (CP) with upper limb disability have difficulty operating standard computer input devices such as, a mouse and a keyboard. The present study evaluated the performance of unimpaired individuals and those with CP in the use of a zero-order spring-centered position joystick and a zero-order isometric joystick when interacting with the computer. We hypothesize that the isometric device due to its stiff resistance would provide better control on cursor movement than the position joystick. The subjects acquired differently sized targets at different distances by positioning a cursor on the target. Performance with the position joystick was superior to that of the isometric joystick. The time to acquire the target was directly proportional to the cursor-target distance A, and indirectly proportional to the target size W. Subjects chose to move more slowly toward the smaller and closer targets and they increased speed for larger and more distant targets. The phase-plane diagram that plots values of the velocity profile of the cursor over its displacement revealed the presence of one large amplitude movement that accounts for the peak velocity of the cursor, and several submovements. Fitts' index of difficulty, log(e)(2W/A) was found to be a good predictor of the movement time in a cursor positioning task for both, unimpaired individuals and those with CP.
 
Article
Actigraphy, the long-term assessment of wrist movements, is used in several research fields, among which are included sleep and circadian rhythms. Actigraphs record movements using accelerometers. The present paper addresses some basic problems and their solutions in the actigraphic assessment of movement, motor symptoms, circadian rest-activity rhythms, and nocturnal agitation in healthy elderly and elderly suffering from a neurodegenerative disease (i.e., Parkinson's disease or Alzheimer's disease) and summarizes the results of previous and ongoing research. First, we have investigated how to filter the accelerometer signal in order to minimize the contribution of accelerations induced by positional changes in the gravitational field--a strong source of artefacts. A bandpass filter from 0.5 to 11 Hz appropriately assesses movement induced accelerations while minimizing gravitational artefact. The application of a bandpass filter from 0.25 to 2 or 3 Hz, as is used in some of the commercially available actigraphs, results in artefacts and moreover biases the slower part of the movement spectrum. It is therefore far from optimal for research on aging, which is associated with a generalized motor slowing. Second, we have proposed an alternative to traditional methods of signal processing in actigraphy, in order to assess both the duration and intensity of movements, and in order to distinguish Parkinsonian tremor. Based on this algorithm, new types of actigraphs have been designed. Third, we have proposed sensitive variables in order to quantify rest-activity rhythm disturbances in healthy elderly subjects and Alzheimer patients, who often present with symptoms of nocturnal restlessness. Since, in these subjects, research protocols applying enforced phase shifts or time-free environments are unfeasible and not justifiable from an ethical point of view, the variables were specifically designed to assess the functionality of the circadian timing system from actigraphic recordings made in the natural environment of subjects. Examples of the application of actigraphy are given, including studies on symptom fluctuations and medication responses in Parkinson patients, and studies on circadian rhythm disturbances and possible remedies in elderly and Alzheimer patients.
 
Article
A barrier to performing more in-depth analyzes during the wheelchair design process is a lack of dynamic reaction force and moment data, and the instrumentation to collect this data. Instrumentation was developed to collect the dynamic force and moment data. New data collections methodologies and analysis techniques were implemented to facilitate computer-aided-engineering for wheelchair designs. Data were collected during standardized wheelchair fatigue tests, while driving over a simulated road course within a laboratory, and while driving in the community. Seventeen subjects participated in this study. Based upon the three test conditions, a pseudo-statistical distribution of the force and moment data at both a caster and rear wheel was developed. The key parameters describing the distribution and the extremums of the data (minima and maxima) were compared using analysis of variance. The results showed that the force and moment distributions and extreme values were similar for the both sets of human trials (i.e., simulated road course and field trials). However, the standardized testing (i.e., wheelchair fatigue testing) differed from both human trials. The force/moment data gathered during this study are suitable for inputs in finite element analysis and dynamic modeling. Our results suggest that the fatigue tests should be modified to change the magnitude and increase the frequency of the forces and moments imparted on the wheelchair. The data reported from this study can be used to improve wheelchair standards and to facilitate computer-aided-engineer in wheelchair design.
 
Article
Selective activation of muscle groups in the feline hindlimb by electrical stimulation of the ventral lumbo-sacral spinal cord was investigated. Spinal cord segments L5 to S1 were mapped using a penetrating tungsten needle electrode. Locations that produced isolated contraction of quadriceps, tibialis anterior or triceps surae/plantaris muscles when stimulated with a current of 40 microA or less, and in which spread of activity to other muscles was not detected after increasing the stimulus to at least twice the threshold level, were defined as belonging to the target muscle's "activation pool." The quadriceps activation pool was found to extend from the beginning of L5 to the middle of L6. The tibialis anterior activation pool extended from the beginning of L6 to the middle of L7, and the triceps surae/plantaris activation pool extended from the caudal end of L6 to the beginning of S1. The three activation pools were located in Rexed motor lamina IX and their spatial organization was found to correspond well with that of the anatomically defined motor pools innervating the same muscles. The spatial and functional segregation of motor pools manifested at the spinal cord level can have direct applications in the areas of functional electrical stimulation and motor control.
 
Scalp maps and mean power spectra of four independent components of the EEG of one subject in 554 3-s epochs beginning 1 s before target stimulus presentations. The ordinate gives root mean square power in all 31 scalp channels in relative log dB units. In the scalp maps, representing the projections to the scalp of the same four independent components, dark and light regions represent projections of opposite polarity. 
Mean target-related event-related spectral perturbation (ERSP) plots for the same four components. The abscissa shows time relative to target presentation; the ordinate, EEG frequency. Times of stimulus presentation and median reaction time (RT) are indicated with broken lines. The color scale shows increases and decreases in spectral power from the pre-stimulus baseline in dB units. The two components labeled ( lower panels ) showed marked and prolonged suppression at three frequencies (near 11 Hz, 23 Hz and 33 Hz) following median reaction time (RT), plus brief augmentations near 16 Hz near 1000 ms. By contrast, ERSP features for the posterior alpha components ( upper panels ) were much weaker and shorter-lasting. 
Article
The prospect of noninvasive brain-actuated control of computerized screen displays or locomotive devices is of interest to many and of crucial importance to a few 'locked-in' subjects who experience near total motor paralysis while retaining sensory and mental faculties. Currently several groups are attempting to achieve brain-actuated control of screen displays using operant conditioning of particular features of the spontaneous scalp electroencephalogram (EEG) including central mu-rhythms (9-12 Hz). A new EEG decomposition technique, independent component analysis (ICA), appears to be a foundation for new research in the design of systems for detection and operant control of endogenous EEG rhythms to achieve flexible EEG-based communication. ICA separates multichannel EEG data into spatially static and temporally independent components including separate components accounting for posterior alpha rhythms and central mu activities. We demonstrate using data from a visual selective attention task that ICA-derived mu-components can show much stronger spectral reactivity to motor events than activity measures for single scalp channels. ICA decompositions of spontaneous EEG would thus appear to form a natural basis for operant conditioning to achieve efficient and multidimensional brain-actuated control in motor-limited and locked-in subjects.
 
Article
The NavChair Assistive Wheelchair Navigation System [7] is an adaptive shared control system being developed to provide mobility to those individuals who would otherwise find it difficult or impossible to use a power wheelchair due to cognitive, perceptual, or motor impairments. The NavChair provides task-specific navigation assistance to the wheelchair operator in the form of several distinct operating modes, each of which distributes control differently between the wheelchair and the operator. This paper describes the NavChair's mechanism for automatically selecting the most appropriate operating mode based on a combination of the wheelchair's immediate situation and its global location. Results from two experimental evaluations of the adaptation method are presented.
 
The interface of the prototype. Two windows show the world viewed through two different cameras. A selected object (in this case, the computer on the table) is indicated with a red bounding box. To the upper left, the toolbar contains buttons for selecting, moving, rotating, linking, and un-linking objects; moving and rotating segments; measuring distances; and taking photos. Below to the right, the “Object Information” window supports editing of an object's name, size, position, rotation, degrees-of-freedom, etc. To the upper right, the “Library” window supports browsing for, and importing new objects into the world. The window shows a preview image and text information of an object (if present), and a file/folder browser. 
Article
A software prototype to support the planning process for adapting home and work environments for people with physical disabilities was designed and later evaluated. The prototype exploits low-cost three-dimensional (3-D) graphics products in the home computer market. The essential features of the prototype are: interactive rendering with optional hardware acceleration, interactive walk-throughs, direct manipulation tools for moving objects and measuring distances, and import of 3-D-objects from a library. A usability study was conducted, consisting of two test sessions (three weeks apart) and a final interview. The prototype was then tested and evaluated by representatives of future users: five occupational therapist students, and four persons with physical disability, with no previous experience of the prototype. Emphasis in the usability study was placed on the prototype's efficiency and learnability. We found that it is possible to realise a planning tool for environmental adaptations, both regarding usability and technical efficiency. The usability evaluation confirms our findings from previous case studies, regarding the relevance and positive attitude towards this kind of planning tool. Although the prototype was found to be satisfactorily efficient for the basic tasks, the paper presents several suggestions for improvement of future prototype versions.
 
Article
An adaptive feedforward control system has been evaluated for use in functional neuromuscular stimulation (FNS) systems. The control system, which utilizes neural network techniques, was used to generate isometric muscle contractions to track a periodic torque trajectory signal. The evaluation of the control system was performed using percutaneous intramuscular electrodes to stimulate the quadriceps muscles of spinal cord injured adolescents. Results of the evaluation indicate that the control system automatically customized its parameters for controlling isometric muscle torque in a particular muscle and that the parameters were adapted on-line to account for changes in muscle properties due to fatigue. This study demonstrates that this control system may play an important role in the development of practical FNS systems that are capable of automatically adjusting stimulation parameters to fit the needs of a particular individual at a given time.
 
Article
A controller was designed for the selective stimulation of the sciatic nerve with a multiple contact cuff electrode to generate a desired torque in the ankle joint of cat. The design integrates three approaches, artificial neural network (ANN) modeling, fuzzy logical adaptation, and geometrical mapping. The geometrical mapping refers to the vector transformation from the joint coordinates to the virtual muscle coordinates which have been conceptually developed to represent the major recruitment features of contact-based functional units in the physical plant. This method reduces the complexity of generating a data set for training the neural network in the feedforward path and implementing the on-line learning algorithm embedded in the feedback loop. The controller was evaluated by computer simulation with the experimental data obtained from the torque generation in five acute cats. The results show that the ANN-based feedforward is capable of predicting 65% of a given desired isometric torque, and the fuzzy logical machine is able to provide suitable gains for feedback modulation to reduce the error from 35 to 8.5% and produce a robust control.
 
Timing of one trial of the experiment with feedback. The cue stimulus in form of an arrow indicates the side of imagination. 
DSLVQ relevance values for 1 Hz frequency components between 9 and 28 Hz (average values for both EEG channels). For subject f3, highest relevance of two frequency bands (9-13 Hz and 21-26 Hz) can be seen; for subject f5 slightly different bands (10-12 Hz and 21-23 Hz) are most relevant. For subject f7, a broad beta band from 18 to 26 Hz shows highest relevance; less pronounced is a narrow band around 12 Hz. For subject g3 a wide frequency band centered at 15 Hz is of highest relevance (13-19 Hz). 
Time courses of the error rate (a) and the averaged time-varying distance (b) for one selected session of all four subjects. The error rates LDA10 (LDA) were obtained by linear discriminant analysis with (without) ten times-ten-fold cross-validation for every 1/8 s. The AAR(6) parameters of both EEG channels were used as features. (b) depicts the ensemble averaged TSD for left and right trials. The TSD can be denoted as the "feature of interest" and is calculated by weighting the AAR parameters with the most discriminating weight vector. 
Article
Electroencephalogram (EEG) recordings during right and left motor imagery can be used to move a cursor to a target on a computer screen. Such an EEG-based brain-computer interface (BCI) can provide a new communication channel to replace an impaired motor function. It can be used by, e.g., patients with amyotrophic lateral sclerosis (ALS) to develop a simple binary response in order to reply to specific questions. Four subjects participated in a series of on-line sessions with an EEG-based cursor control. The EEG was recorded from electrodes overlying sensory-motor areas during left and right motor imagery. The EEG signals were analyzed in subject-specific frequency bands and classified on-line by a neural network. The network output was used as a feedback signal. The on-line error (100%-perfect classification) was between 10.0 and 38.1%. In addition, the single-trial data were also analyzed off-line by using an adaptive autoregressive (AAR) model of order 6. With a linear discriminant analysis the estimated parameters for left and right motor imagery were separated. The error rate obtained varied between 5.8 and 32.8% and was, on average, better than the on-line results. By using the AAR-model for on-line classification an improvement in the error rate can be expected, however, with a classification delay around 1 s.
 
Article
Row-column scanning is a very slow method of communication. Options for increasing text entry rate include 1) dynamically changing the configuration of the row-column matrix or 2) using rate-enhancement techniques like word prediction, but evidence suggests that increased cognitive load imposed by these methods on the user can result in little or no improvement in text generation rate. An alternative we are investigating is adapting a system's scan delay during run-time. Our goal is to allow a scanning system to adjust its parameters "on the fly" (as opposed to the current practice of setting parameters during clinical assessments). This paper describes the evolution of a one-switch row-column scanning system that adapts its scan rate based on measurements of user performance. Two experiments have been performed to explore the effects of automatically adapting scan delay on users' text entry rate. Our results indicate that automatic adaptation has the potential to enhance text-entry rate without increasing task complexity.
 
Article
Using computer simulation, the theoretical feasibility of functional electrical stimulation (FES) assisted standing up is demonstrated using a closed-loop self-adaptive fuzzy logic controller based on reinforcement machine learning (FLC-RL). The control goal was to minimize upper limb forces and the terminal velocity of the knee joint. The reinforcement learning (RL) technique was extended to multicontroller problems in continuous state and action spaces. The validated algorithms were used to synthesize FES controllers for the knee and hip joints in simulated paraplegic standing up. The FLC-RL controller was able to achieve the maneuver with only 22% of the upper limb force required to stand-up without FES and to simultaneously reduce the terminal velocity of the knee joint close to zero. The FLC-RL controller demonstrated, as expected, the closed loop fuzzy logic control and on-line self-adaptation capability of the RL was able to accommodate for simulated disturbances due to voluntary arm forces, FES induced muscle fatigue and anthropometric differences between individuals. A method of incorporating a priori heuristic rule based knowledge is described that could reduce the number of the learning trials required to establish a usable control strategy. We also discuss how such heuristics may also be incorporated into the initial FLC-RL controller to ensure safe operation from the onset.
 
Article
With one or two switches, Morse code could provide an effective alternative communication channel for individuals with physical limitations. However, most of the physically disabled persons have difficulties in maintaining a stable typing of Morse code, and hence the automated recognition of unstable Morse code is becoming more on demand. In this study, an adaptive variable-ratio threshold prediction (AVRTP) algorithm is proposed to analyze the Morse code time series with variable unit time period and ratio. Two least-mean-square (LMS) predictors are applied to track the dot interval and the dot-dash difference concurrently, and then a predicted threshold based on a variable-ratio decision rule is used to distinguish between dots and dashes. The same method is also applied to identify character-spaces. By the adaptive prediction of variable-ratio threshold, AVRTP has successfully overcome the difficulty of analyzing severely unstable Morse code time series and outperformed the previously proposed adaptive unstable-speed prediction (AUSP) algorithm and LMS and matching (I,MS&M) algorithm. This study concludes with a computer simulation and a preliminary clinical evaluation that demonstrate AVRTP as an efficient and reliable method for unstable Morse code recognition.
 
Article
This study quantitatively examined and compared the shoulder motions of C4 level spinal cord injury (SCI), C5 level SCI, and able-bodied persons as a command source. The study was motivated by both the success of shoulder control in functional electrical stimulation (FES) systems designed for C5 level SCI people and the lack of quantitative information on the shoulder motion of persons with C4 level SCI. A dual-axis transducer was used to monitor the elevation/depression and protraction/retraction angles of each subject's shoulder while they performed three experimental sections which examined: the range of active shoulder motion; the ability to move incrementally to discrete positions with the aid of visual feedback; and the ability to hold discrete shoulder positions for an extended period without visual feedback. Results indicated that each group had the largest average shoulder displacements (abled = 23 degrees +/- 4 degrees, C5's = 14 degrees +/- 3 degrees, and C4's = 9 degrees +/- 3 degrees) while attempting to elevate and that on average the C4 group had the smallest range of active shoulder motion. No statistically significant differences between the groups were found in either the accuracy or stability of reaching discrete positions with the aid of visual feedback or in the accuracy of holding discrete shoulder positions for an extended period without visual feedback. The results suggest that within their limited range of motion the individuals with C4 level SCI retained shoulder control sufficient for use as an neuroprosthetic command interface.
 
Article
A portable sound processor has been developed to facilitate research on advanced hearing aids. Because it is based on a digital signal processing integrated circuit (Motorola DSP56001), it can readily be programmed to execute novel algorithms. Furthermore, the parameters of these algorithms can be adjusted quickly and easily to suit the specific hearing characteristics of users. In the processor, microphone signals are digitized to a precision of 12 bits at a sampling rate of approximately 12 kHz for input to the DSP device. Subsequently, processed samples are delivered to the earphone by a novel, fully-digital class-D driver. This driver provides the advantages of a conventional class-D amplifier (high maximum output, low power consumption, low distortion) without some of the disadvantages (such as the need for precise analog circuitry). In addition, a cochlear implant driver is provided so that the processor is suitable for hearing-impaired people who use an implant and an acoustic hearing aid together. To reduce the computational demands on the DSP device, and therefore the power consumption, a running spectral analysis of incoming signals is provided by a custom-designed switched-capacitor integrated circuit incorporating 20 bandpass filters. The complete processor is pocket-sized and powered by batteries. An example is described of its use in providing frequency-shaped amplification for aid users with severe hearing impairment. Speech perception tests confirmed that the processor performed significantly better than the subjects' own hearing aids, probably because the digital filter provided a frequency response generally closer to the optimum for each user than the simpler analog aids.
 
Article
An electronic circuit to convert the information contained in a digital display to natural voice is presented. Extraction of the information is done on the display itself using buffered inputs to the converter. It is changed to voice by a microcontroller and a voice chip.
 
Article
Several in-the-ear (ITE) and behind-the-ear (BTE) hearing aids were tested for audible interference at various distances from five types of digital wireless telephones. The interference which takes the form of a buzzing and a static sound was quantified using a calibrated system including a frequency analyzer and a pressure field microphone. The output of the each hearing aid was coupled to the microphone via Tygon tubing and a standard 2 cc coupler. The highest interference-induced sound pressure level (SPL), 122.5 dB, was measured from a BTE hearing aid placed within 2 cm of a transmitting Global System for Mobile Communications (GSM) phone. In this case, interference was detected up to a separation distance of almost 3 m. While all phones tested produced a similar interference level within 2 cm of this hearing aid, interference SPL from the code division multiple access (CDMA)-based system decreased more rapidly with distance than the time division multiple access (TDMA)-based phones tested.
 
Article
Functional electrical stimulation (FES) is a means of restoring gait to individuals with spinal cord injury, but the performance of most FES-aided gait systems is hampered by the rapid muscle fatigue which results from stimulated muscle contraction and the inadequate control of joint torques necessary to produce desired limb trajectories. The controlled-brake orthosis (CBO) addresses these limitations by utilizing FES in combination with a long-leg brace that contains controllable friction brakes at the knees and hips. A laboratory version of the CBO utilizing computer-controlled magnetic particle brakes at the joints was designed and constructed, and preliminary results with a single spinal cord injury (SCI) subject have demonstrated reduced fatigue and more repeatable gait trajectories when compared to FES-aided gait without the brace. Significant work remains to demonstrate the efficacy of the concept across a wide range of SCI subjects and to design a system which meets appropriate user requirements of size, weight, cosmesis, ease of use and cost. The primary purpose of the paper is to detail the design of the CBO.
 
Article
Our goal is to apply robotics and automation technology to assist, enhance, quantify, and document neurorehabilitation. This paper reviews a clinical trial involving 20 stroke patients with a prototype robot-aided rehabilitation facility developed at the Massachusetts Institute of Technology, Cambridge, (MIT) and tested at Burke Rehabilitation Hospital, White Plains, NY. It also presents our approach to analyze kinematic data collected in the robot-aided assessment procedure. In particular, we present evidence 1) that robot-aided therapy does not have adverse effects, 2) that patients tolerate the procedure, and 3) that peripheral manipulation of the impaired limb may influence brain recovery. These results are based on standard clinical assessment procedures. We also present one approach using kinematic data in a robot-aided assessment procedure.
 
Article
Scaphoid or longitudinal arch pads are frequently prescribed pedorthics for foot and ankle rehabilitation. These pedorthics are reported to be effective in mechanically supporting the medial longitudinal arch while reducing plantar and medial soft tissue strain. The objective of this study was to measure alterations in ambulatory plantar pressure metrics in a group of adults secondary to scaphoid pad application. The biomechanical rationale of this study was that the geometry of foot contact would be altered secondary to foot inversion. Ten adult male subjects with biomechanically normal feet were evaluated during multiple trials. A Holter type microprocessor-based portable in-shoe plantar pressure data acquisition system was used to record the dynamic data. Pressures were recorded from eight discrete plantar locations at the hindfoot, midfoot, and forefoot regions of the insole. Statistically significant (p < or = 0.05) increases in peak pressures were seen laterally with scaphoid pad application, while significant decreases in peak pressures with pad usage occurred at the hallux and the calcaneal region of the foot. At the medial longitudinal arch, peak pressures increased from near 0 to 115.3 kPa, contact durations increased from near 0 to 438 ms, and pressure-time integrals increased from near 0 to 33.4 kPa.s.
 
Ambiguous keyboard arrangements proposed by earlier researchers. The space character, represented by an underscore, is encoded unambiguously on its own key. 
The performance of keyboards optimized for four diierent prediction algorithms, as compared to four earlier keyboard arrangements. The average keystroke eeciencies and the eeciencies for the individual testing texts are indicated. 
Article
Many persons with disabilities lack the fine motor coordination necessary to operate traditional keyboards. For these individuals, ambiguous (or reduced) keyboards offer an alternative access method. By placing multiple characters on each key, the size and accessibility of the individual keys can be enhanced without requiring a larger keyboard. Using statistical disambiguation algorithms to automatically interpret each keystroke, these systems can approach the keystroke efficiency (keystrokes per character) of conventional keyboards. Since the placement of characters on each key determines the effectiveness of these algorithms, several methods of optimizing keyboard arrangements have previously been proposed. This paper presents a new method for optimizing an arbitrary set of N characters over a collection of M keys. While earlier efforts relied upon approximations of keystroke efficiency, the proposed approach optimizes the arrangement under this exact performance measure. Applied to the canonical 26 characters on a nine-key "telephone keypad" problem, this method provides an improvement in efficiency of 2.5 percentage points over previously established layouts. Using only a minimum of calculations, the proposed technique operates quickly and efficiently, deriving optimal arrangements in a matter of seconds using a personal computer. The flexible method is applicable to arbitrary disambiguation algorithms, character sets, and languages.
 
Article
A minimal set of muscles (8 to 16) were identified as candidates for implantation in a clinical system to provide walking function to individuals with complete paraplegia using functional electrical stimulation (FES). Three subjects with complete motor and sensory paraplegia had percutaneous intramuscular electrodes implanted in all major muscles controlling the trunk, hips, knees, and ankles. Stimulation patterns for walking with FES were generated for different sets of eight and 16 muscles. The quality and repeatability of the resulting gait produced by walking patterns consisting of various combinations of muscles were determined. Most eight-channel stimulation patterns resulted in scissoring or insufficient hip flexion, preventing forward progression. One eight-channel system allowed a maximum speed of 0.1 m/s with a cadence of 22 steps/min and a stride length less than 0.3 m. Improved walking performance was observed with 16 channels of stimulation. This ranged from slow step- to gait at 0.1 m/s to smooth reciprocal gait at 0.5 m/s. In all three subjects, the favored combination of 16 channels included erector spinae for trunk extension; gluteus maximus, posterior portion of adductor magnus and hamstrings for hip extension; tensor fasciae latae and either sartorius or iliopsoas for hip flexion; vastus lateralis/intermedius for knee extension; and tibialis anterior/peroneous longus for ankle dorsiflexion. In one subject the 16-channel FES system provided repeatable day-to-day gait averaging 0.4 m/s, 58 steps/min and a stride length at 0.8 m. A maximum repeatable walking distance with 16 channels was 34 m. Multiple 34-m trials were possible with minimal rests between walks. Fatigue of both the hip extensors and upper body was a limiting factor. The selection of target muscles for implantation is critical to the performance of FES systems. This study provides guidelines to muscle selection for walking with FES based on objective measures of gait performance. The findings indicate that a 16-channel FES system for total implantation is feasible for repeatable short distance, independent, walker-support walking in paraplegia.
 
Foot velocity patterns show group (N = 9) mean curves for the nonamputee controls (thick solid line) and transtibial amputees wearing the Single Axis (dotted line), Seattle Lite (thin solid line), and Flex Foot (dashed line). 
Shank velocity patterns show group (N = 9) mean curves for the nonamputee controls (thick solid line) and transtibial amputees wearing the Single Axis (dotted line), Seattle Lite (thin solid line), and Flex Foot (dashed line). 
Thigh velocity patterns show group (N = 9) mean curves for the nonamputee controls (thick solid line) and transtibial amputees wearing the Single Axis (dotted line), Seattle Lite (thin solid line), and Flex Foot (dashed line). 
Intersegmental mobility at the ankle joint in nonamputee (NA) controls and transtibial amputees wearing the SA, SL, and FF foot types shown by mean foot velocity (solid lines) and shank velocity (dashed lines). The solid arrows indicate transitions in prosthetic ankle motion from plantarflexion to dorsiflexion, while the dashed arrow represents a much earlier transition in the control group. 
Intersegmental mobility at the knee joint in nonamputee (NA) controls and transtibial amputees (TTA) wearing the SA, SL, and FF foot types shown by mean thigh velocity (solid lines) and shank velocity (dashed lines). The solid arrows indicate transitions in TTA knee motion from flexion to extension, while the dashed arrow represents a much earlier transition in the control group.
Article
Dynamic elastic response foot and ankle prostheses (Seattle-Lite, Flex Foot, etc.) used by transtibial amputees feature substantial design improvements over conventional designs (SACH, Single Axis, etc.). Despite this progress, transtibial amputees continue to expend greater energy than normals. Increased residual limb EMG data and altered gait patterns suggest that impaired mobility may be the cause of overactive muscles in early stance. Prosthetic mobility was therefore quantified by measuring foot, shank and thigh velocities in nine transtibial amputees, wearing three different foot designs: Single Axis (SA), Seattle Lite (SL) and Flex Foot (FF). The magnitude, timing and rate of segment velocities for each prosthetic design, characterizing early stance mobility, were compared with corresponding measures in normal, nonamputee (NA) controls using Dunnett's test. Regardless of foot type, transtibial (TT) amputees walked slower than non amputee controls (63.3-65.8 m/min versus 78.5 m/min, p < 0.05) and their stride length was shorter (1.21-1.26 m versus 1.41 m, p < 0.01). In early stance, peak foot and shank velocities were lower (p < 0.01) for both the SL and FF while only shank velocity was lower (p < 0.01) with the SA compared to NA controls. Significant delays in the timing of early stance events such as peak shank velocity, peak ankle plantarflexion and peak knee flexion compromised shank and knee stability in TT amputees. Foot and shank mobility was uncontrolled with the SA design while ankle mobility was restricted by the FF and SL feet. In NA controls on the other hand, appropriate timing and rate of segment velocity changes preserved dynamic stability and forward progression in early stance. This was evidenced by rapid decreases in foot and shank velocity as the thigh velocity increased during weight acceptance. Future prosthetic designs should provide TT amputees with improved ankle mobility that attempt to capture the dynamic characteristics of a normal articulation between the foot and shank segments during the early stance weight acceptance period.
 
Times for the calculation of ratios and quotients. 
The key transition points in one cycle PGG signals. 
(a) The overlap when 1 = 2 , only one intersection point. (b) The overlap when 1 < 2 , two intersection points. 
(a) PGG and EGG waveforms from normal subject DS11. (b) PGG and EGG waveforms from subject JV01 with recurrent laryngeal paralysis. (c) PGG and EGG waveforms from subject BR21 with superior laryngeal paralysis. 
Article
Electroglottography (EGG) and photoglottography (PGG) are two plausible methods to study voice production for monitoring the patterns of laryngeal vibrations. It has been suggested that measures such as open quotient and speed quotient calculated from glottographic signals can provide useful information regarding pathological phonation. In this paper, an integrated analyzer and classifier of glottographic signals was implemented. The system makes it possible to calculate the measures from digitized EGG and PGG signals automatically in order to examine vocal fold abnormality. The system developed several techniques to extract features from glottographic signals and proposed a statistical classification method that can possibly aid the diagnosis process. To check the reliability of the system, a training set and a test set of glottographic signals from normal people and patients with recurrent/superior laryngeal paralysis were analyzed and classified by the system. The results showed that the system is a useful tool for quantitative study of phonatory pathophysiology and can be used by the examiner who is interested in the clinical examination of glottographic signals. Moreover, glottographic techniques may have some clinical applications in the quantitative documentation of phonatory function in patients with voice disorders but requires further evaluation before clinical application.
 
Tracing of a hemiplegic hand with active electrode (cathode) on dorsal surface of forearm, indifferent electrode (anode) on ventral surface close to wrist crease and motion sensors on fingertips. Wrist at 30 flexion. (a) before stimulation and (b) during stimulation. 
Hand aperture versus wrist angle; pooled data from all five subjects. Open circles: distances between digits without stimulation, filled circles: distances between tips of digits during hand-opening stimulation. (a) thumb to forefinger (digit 1 to digit 2: D1-D2), (b) thumb to digit 3, and (c) thumb to digit 4 or 5, whichever distance was smaller during stimulation. Thick lines: second-order linear regression curves, thin lines: 95th percentile confidence intervals. 
Article
This paper studied the effect of wrist angle on the amount of hand opening achieved by electrical stimulation in people with spastic hemiplegia. With their forearm in pronation, subjects were asked to relax while their affected wrist was passively moved in steps of about 15 degrees from full flexion into extension. Trains of stimuli were applied to the long finger extensor muscles through surface electrodes on the forearm. At each wrist position stimulation was turned on for a few seconds until hand opening equilibrated. Wrist angle and fingertip positions were recorded using a three-dimensional (3-D) motion analysis system. Maximal displacements between thumbtip and each fingertip occurred when the wrist was fully flexed. As the wrist was extended, hand aperture achieved by electrical stimulation progressively declined, reaching zero at 40 degrees of wrist extension. We conclude that electrical stimulation can significantly increase the grasp aperture of the hemiplegic hand, but this is strongly dependent on wrist posture and accompanying voluntary effort.
 
Article
This study presents a possible solution of the general problem of coordinating muscle stimulation in a neuroprosthesis when multiarticular muscles introduce mechanical coupling between joints. In a hand-grasp neuroprosthesis, extrinsic hand muscles cross the wrist joint and introduce large wrist flexion moments during grasp. In order to control hand grasp and wrist angle independently, a controller must take the mechanical coupling into account. In simulation, we investigated the use of artificial neural networks to coordinate hand and wrist muscle stimulation. The networks were trained with data that is easily obtained experimentally. Feedforward control showed excellent hand and wrist coordination when the properties of the system were fixed and there were known external loads. Predictable disturbances (e.g., gravity acting on the hand) can be compensated by sensing arm orientation. However, since wrist angle is sensitive to unpredictable disturbances (e.g., fatigue or object weight), voluntary intervention or feedback control may be required to reduce residual errors.
 
Article
The purpose of this study was to evaluate a "suture" type electrode for direct bladder stimulation in an animal model of a lower motor neuron lesion. During an initial surgery, five male cats were instrumented under anesthesia using multistranded, 316 LVM, stainless-steel, wire electrodes implanted on the bladder wall serosa above the trigone area. Electrodes were constructed with a needle attached to the end that was removed after suturing the electrode in place. Additional instrumentation included urinary bladder catheters (tubes) for pressure recording and filling, and hook type electrodes for leg and pelvic floor electromyography recording. Chronic bladder filling and stimulation studies were conducted in tethered animals three to four weeks following surgery. To test these electrodes in a spinal cord injury model, a lower motor neuron lesion was performed including the sacral cord and complete nerve roots at L6 and below. These animals were evaluated during weeks 3 and 10 after injury. Direct bladder stimulation induced active contractions and voiding both before and after spinal cord injury. Effective stimulation parameters consisted of 40 pulses per s, 300 micros to 1 ms pulse duration, a stimulation period from 3 to 4 s, and a stimulation current from 10 to 40 mA. Fluoroscopy revealed an open membranous urethra during stimulation and following stimulation. A small diameter penile urethra was observed to limit flow. Postmortem evaluation of the suture electrode revealed no abnormalities such as corrosion, migration into the bladder lumen or displacement. These findings indicate that suture electrodes are suitable and effective for short-term implantation in the lower motor neuron animal model.
 
Article
A five-camera Vicon (Oxford Metrics, Oxford, England) motion analysis system was used to acquire foot and ankle motion data. Static resolution and accuracy were computed as 0.86 +/- 0.13 mm and 98.9%, while dynamic resolution and accuracy were 0.1 +/- 0.89 and 99.4% (sagittal plane). Spectral analysis revealed high frequency noise and the need for a filter (6 Hz Butterworth low-pass) as used in similar clinical situations. A four-segment rigid body model of the foot and ankle was developed. The four rigid body foot model segments were 1) tibia and fibula, 2) calcaneus, talus, and navicular, 3) cuneiforms, cuboid, and metatarsals, and 4) hallux. The Euler method for describing relative foot and ankle segment orientation was utilized in order to maintain accuracy and ease of clinical application. Kinematic data from a single test subject are presented.
 
Article
In this paper, we report on our pilot evaluation of a prototype foot/ankle prosthesis. This prototype has been designed and fabricated with the intention of providing decreased ankle joint stiffness during the middle portion of the stance phase of gait, and increased (i.e., more normal) knee range of motion during stance. Our evaluation involved fitting the existing prototype foot/ankle prosthesis, as well as a traditional solid ankle cushioned heel (SACH) foot, to an otherwise healthy volunteer with a below-knee (BK) amputation. We measured this individual's lower extremity joint kinematics and kinetics during walking using a video motion analysis system and force platform. These measurements permitted direct comparison of prosthetic ankle joint stiffness and involved side knee joint motion, as well as prosthetic ankle joint moment and power.
 
Article
Previous studies investigating postural control using platform perturbations demonstrated that forward leaning occurs under certain experimental conditions. This study examined a potential benefit of forward leaning, investigating the hypothesis that forward lean acts to increase the effective stiffness of the postural ankle dynamics. A systems modeling approach was used to evaluate the effect of forward lean (4 degrees ankle flexion beyond normal stance) on dynamic postural responses to continuous random antero-posterior platform acceleration in four healthy young adults. Motion was limited to the ankle joint using a restraint device and responses were characterized in terms of center-of-pressure displacement. Also, EMG and kinematic data were used to partition measured ankle torque into impedance and activation (postural "reflex") components. The results failed to show a consistent effect, due to lean, on the ankle dynamics. A significant increase in phasic ankle torque due to increased muscle impedance during forward lean was counterbalanced by a comparable decrease in net "reflex" ankle torque. The change in "reflex" torque was apparently due to decreased phasic dorsiflexor activity in forward lean, since phasic plantarflexor activity did not change significantly. Functionally, the results suggest that forward lean occurs for reasons other than stiffening of the overall postural ankle dynamics.
 
Article
Anterior knee pain (AKP) is a common pathological condition, particularly among young people and athletes, associated to an abnormal motion of the patella during the bending of the knee and possibly dependent on a muscular or structural imbalance. A lack of synergy in the quadriceps muscles results in a dynamic misalignment of the patella, which in turn produces pain. AKP rehabilitative therapy consists of conservative treatment whose main objective is to strengthen the Vastus Medialis. The aim of this article is to study the quadriceps muscle control strategy in AKP patients during an isokinetic exercise. Analysis of the muscle activation strategy is important for an objective measurement of the knee functionality in that it helps to diagnose and monitor the rehabilitative treatment. Surface electromyography (EMG) from the three superficial muscles of the femoral quadriceps during a concentric isokinetic exercise has been analyzed along with the signals of knee joint position and torque. A group of 12 AKP patients has been compared with a group of 30 normal subjects. Analysis of the grand ensemble average of the EMG linear envelopes in AKP patients reveals significant modifications in Vastus Medialis activity compared to the other quadriceps muscles. In order to study the synergy of the muscles, temporal identifiers have been associated to the EMG linear envelopes. To this end, EMG linear envelope decomposition in Gaussian pulses turned out to be effective and the results highlight an appreciable delay in the activation of the Vastus Medialis in AKP patients. This muscular unbalance can explain the abnormal motion of the patella.
 
Article
Spatiotemporal arm and body movements of able-bodied subjects performing nine everyday tasks were recorded for the purpose of guiding the development of an upper-limb orthosis. To provide a user the opportunity to carry out these tasks with natural movements, the orthosis should allow replication of the measured trajectories. We outline the orthosis architecture, which supports the user's upper arm and forearm, and analyze the movement data to obtain orthosis design specifications. Trajectories were obtained using six-degree-of-freedom magnetic position sensors affixed to the wrist, elbow, shoulder, trunk and head. Elbow trajectory data were decomposed into ranges along the principle Cartesian axes to provide a generally useful envelope measure. The smallest Cartesian parallel-piped that contained the elbow trajectories for most tasks was approximately 30 cm front/back, 15 cm side/side, and 17 cm up/down. A rough lower bound estimate obtained by asking subjects to repeat the tasks while minimizing elbow movement substantially reduced movement in the up/down and side/side dimensions. Elbow angles were generally in the range 50 degrees-150 degrees, and the angle of the forearm with respect to vertical was 10 degrees-110 degrees. Raw trajectory data may be downloaded from www://asel.udel.edu/robotics/orthosis/range.h tml.
 
Article
In an effort to assess the safety and efficacy of focal intracortical microstimulation (ICMS) of cerebral cortex with an array of penetrating electrodes as might be applied to a neuroprosthetic device to aid the deaf or blind, we have chronically implanted three trained cats in primary auditory cortex with the 100-electrode Utah Intracortical Electrode Array (UIEA). Eleven of the 100 electrodes were hard-wired to a percutaneous connector for chronic access. Prior to implant, cats were trained to "lever-press" in response to pure tone auditory stimulation. After implant, this behavior was transferred to "lever-presses" in response to current injections via single electrodes of the implanted arrays. Psychometric function curves relating injected charge level to the probability of response were obtained for stimulation of 22 separate electrodes in the three implanted cats. The average threshold charge/phase required for electrical stimulus detection in each cat was, 8.5, 8.6, and 11.6 nC/phase respectively, with a maximum charge/phase of 26 nC/phase and a minimum of 1.5 nC/phase thresholds were tracked for varying time intervals, and seven electrodes from two cats were tracked for up to 100 days. Electrodes were stimulated for no more than a few minutes each day. Neural recordings taken from the same electrodes before and after multiple electrical stimulation sessions were very similar in signal/noise ratio and in the number of recordable units, suggesting that the range of electrical stimulation levels used did not damage neurons in the vicinity of the electrodes. Although a few early implants failed, we conclude that ICMS of cerebral cortex to evoke a behavioral response can be achieved with the penetrating UIEA. Further experiments in support of a sensory cortical prosthesis based on ICMS are warranted.
 
Article
Animal models for joint diseases are necessary for in vivo studies. Joint contractures are characterized by lack of the normal range of motion of a joint most often due to increased soft tissue stiffness. Biological and biochemical data have been obtained but biomechanical data on small animals are rare. An instrument was developed to measure rat knee angular displacement at various soft tissue loads in normal and pathological circumstances. This article describes the instrument and reports its reproducibility and accuracy. The reproducibility and accuracy of this instrument was found to be acceptable thereby validating its use for research purposes with adult rat knees.
 
Article
This study describes a new method for evaluating polyethylene wear in total knee arthroplasty. Since the amount of wear is dependent on a number of variables such as the weight and activity of the patient, it should be estimated based on in vivo measurements. We used a computer vision technique called three-dimensional/two-dimensional (3-D/2-D) matching to perform in vivo assessment using a single-plane radiograph. Using the 3-D/2-D matching algorithm we estimated the 3-D position and orientation of each knee implant and then measured the femorotibial distance, which is defined as the shortest perpendicular distance from the tibial tray to the femoral component. The accuracy of the proposed 3-D/2-D matching method was determined by in vitro investigations. The worst errors in in-plane/out-of-plane translations and rotations were 0.20/1.95 mm and 0.17/0.29 degrees, respectively. The root-mean-square error in femorotibial distance measurements using real polyethylene inserts was 0.04 mm. Results of in vivo femorotibial distance measurements are also described.
 
Article
This paper presents an electromyographic (EMG) pattern recognition method to identify motion commands for the control of a prosthetic arm by evidence accumulation based on artificial intelligence with multiple parameters. The integral absolute value, variance, autoregressive (AR) model coefficients, linear cepstrum coefficients, and adaptive cepstrum vector are extracted as feature parameters from several time segments of EMG signals. Pattern recognition is carried out through the evidence accumulation procedure using the distances measured with reference parameters. A fuzzy mapping function is designed to transform the distances for the application of the evidence accumulation method. Results are presented to support the feasibility of the suggested approach for EMG pattern recognition.
 
The percent of correct bootstrapped trials obtained at each level of communication speed. Note that the speed of communication is inversely related to the number of trials averaged. 
The data flow, and logical structure, of the on-line, real time, application of the BCI. 
Article
We describe a study designed to assess a brain-computer interface (BCI), originally described by Farwell and Donchin [9] in 1988. The system utilizes the fact that the rare events in the oddball paradigm elicit the P300 component of the event-related potential (ERP). The BCI presents the user with a matrix of 6 by 6 cells, each containing one letter of the alphabet. The user focuses attention on the cell containing the letter to be communicated while the rows and the columns of the matrix are intensified. Each intensification is an event in the oddball sequence, the row and the column containing the attended cell are "rare" items and, therefore, only these events elicit a P300. The computer thus detects the transmitted character by determining which row and which column elicited the P300. We report an assessment, using a boot-strapping approach, which indicates that an off line version of the system can communicate at the rate of 7.8 characters a minute and achieve 80% accuracy. The system's performance in real time was also assessed. Our data indicate that a P300-based BCI is feasible and practical. However, these conclusions are based on tests using healthy individuals.
 
Article
This paper describes the results of an exploratory study of the use of standard occupational therapy assessment tests to measure the effective manipulation ability of individuals with disabilities using a robotic aid. Robotic manipulators have been explored for use as a vocational accommodation to support the job placement of individuals with severe manipulation disabilities. One of the factors that has impeded the transfer of this work is the lack of practical information that is relevant to the vocational placement process. The preliminary performance data presented in this paper provides an indication of robot-assisted manipulation skill that rehabilitation professionals may use to better understand the potential for use of this technology in providing greater job opportunities for people with severe manipulation impairment. Three different assessment tests were administered to nine different subjects with severe physical disabilities using a computer-controlled robotic workstation to perform the manipulation requirements of the tests. In all cases, the subjects, who were otherwise unable to physically perform the tasks without the robot, were able to perform manipulation tasks a factor of 20-700 times less than that of the performance indicated in published norms. Although these performance levels are modest in terms of nondisabled populations, supporting data is also provided that suggests that individuals with severe manipulation deficits could have access to a much wider range of vocational opportunities with an appropriate implementation of robot technology.
 
Article
To promote proper wheelchair securement in transportation, the proposed ANSI/RESNA Standard on Wheelchairs Used as Seats in Motor Vehicles will require that all transit wheelchairs be equipped with four securement points compatible with strap-type tiedowns. Through computer simulations, the location of these securement points has been found to influence wheelchair user response to a frontal crash. This study develops and employs an injury risk assessment method to compare the crashworthiness of various securement point configurations. The comparative injury risk assessment method is designed to predict the risk associated with internalized crash forces, as well as risk associated with secondary occupant impact with the vehicle interior. Injury criteria established by Federal Motor Vehicle Safety Standards and General Motors, along with excursion limitations set by the Society of Automotive Engineers (SAE) J2249 Wheelchair Tiedowns and Occupant Restraint Systems (WTORS) Standard were used as benchmarks for the risk assessment method. The simulation model subjected a secured commercial powerbase wheelchair with a seated 50th percentile male Hybrid III test dummy to a 20 g/30 mph crash. The occupant was restrained using pelvic and shoulder belts, and the wheelchair was secured with four strap-type tiedowns. Results indicated that securement points located 1.5 in to 2.5 in above the evaluated wheelchair's center of gravity provide the most effective occupant protection.
 
Article
The principle of using robotic techniques to assist an active upper limb exercise is demonstrated in ten patients with weakness and spasticity. Using a servo motor to apply torque about the elbow, the mean range of active extension-flexion was increased in every patient. Sample kinematic and electromyographic (EMG) data are given.
 
Top-cited authors
Gert Pfurtscheller
  • Graz University of Technology
Herbert Ramoser
  • Commend International
Emanuel Donchin
  • University of South Florida
Jonathan Wolpaw
  • Albany Stratton VA Medical Center
Dennis Mcfarland
  • Wadsworth Center, NYS Department of Health