Nonlinear optimisation-based search algorithms have been developed for the precise stimulation of muscles in the wrist and hand, to enable stroke patients to attain predefined gestures. These have been integrated in a system comprising a 40 element surface electrode array that is placed on the forearm, an electrogoniometer and data glove supplying position data from 16 joint angles, and custom signal generation and switching hardware to route the electrical stimulation to individual array elements. The technology will be integrated in a upper limb rehabilitation system currently undergoing clinical trials to increase their ability to perform functional tasks requiring fine hand and finger movement. Initial performance results from unimpaired subjects show the successful reproduction of six reference hand postures using the system.
The aim of this work was to assess the feasibility of a treatment based on volitional cycling augmented by Functional Electrical Stimulation (FES) on hemiparetic adolescents. Six chronic hemiparetic adolescents were included in a case series study. Patients underwent FES-cycling training combined with voluntary pedaling. The intervention consisted of 21 sessions lasting 30 minutes each. Patients were evaluated before, after training, and at a 3-month follow-up visit through clinical scales (Winter scale, observational gait scale, gross motor function measurement, Boyd test and Ashworth scale), a standard gait analysis and a voluntary pedaling test. Results were compared with an age-matched healthy control group (N=6). Two subjects withdrew the study before the completion of the intervention. Concerning the four remaining subjects, the clinical scales showed a slight level of disability already at baseline and no changes were observed after the intervention. In terms of walking ability, some significant improvements (Kruskal-Wallis test, p-value<0.05) were obtained after training in two out of four subjects: an increase of about 16% and 41% of the ankle range of motion and of about 18% and 33% of the ankle propelling power were achieved for two subjects, respectively. During pedaling, the work produced by the paretic leg while pulling the pedal significantly increased in 3 out of 4 subjects. In one subject a more symmetrical cycling movement was observed, whilst for another subject a significant improvement in terms of co-contraction between rectus femoris and biceps femoris was achieved. In conclusion, this study assessed the feasibility of FES-cycling training on hemiparetic adolescents, but did not provide evidences about the effectiveness of this intervention in improving motor recovery and walking ability. However, since only a small group of patients with a low level of disability was involved in the study, further investigations are needed to provide conclusive results.
We present a novel system for control of elbow movements by electrical stimulation of the biceps and triceps in tetraplegic patients. The operation of the system uses the novel algorithm and applies closed loop control. Movement of the arm is generated via multi-pad electrodes developed by Tecnalia Serbia, Ltd. by the stimulator that allows asynchronous activation of individual pads. The electrodes are positioned over the innervation of the biceps and triceps muscles on the upper arm. This layout allows distributed activation; thereby, selective and low fatiguing activation of paralyzed muscles. The sensory feedback comes from the image acquired by the Microsoft Kinect system and the depth stream analysis is performed in real time by the computer running in the MatLab environment. The image based feedback allows control of the hand position at the target by cocontraction of the antagonists. The control adjusts the stimulation intensity and results with the tracking of the desired movement. The algorithm was proven to operate efficiently in a tetraplegic patient.
Muscle fatigue due to functional electrical stimulation still prevents its widespread use as a gait rehabilitation tool for spinal cord injured subjects. Although there is an active research towards optimization of pulse parameters to delay muscle fatigue, changes in stimulated muscle's performance during repeated contractions due to fatigue have not been yet determined. In this work, a study conducted in two phases with a sample of incomplete spinal cord injured patients is presented. In the first phase, a fatigue protocol based on submaximal isometric contractions allowed to obtain an objective criterion for estimation of fatigue of knee muscles from initial changes in muscle performance. The criterion is incorporated in the fatigue protocol in the second phase of the study, to compare two novel customized fatigue management strategies. Results showed that knee flexor muscles develop less force and lower fatigue than extensor muscles. Muscle fatigue management strategies based on customized modulations of stimulation frequency are valid to delay muscle fatigue.
Walking was tested in 4 people with thoracic-level paraplegia using stimulation of quadriceps muscles, the flexor reflex and unlocking knee-ankle-foot orthoses (KAFO). Heart rate, speed, distance, kinematics and ground reaction forces were measured while subjects walked using a walker. None of the subjects could walk without the system; all could walk continuously for at least 4 minutes with it. Joint angles and some other kinematic features resembled normal walking, but the walking was too slow (average speed: 3.8 m/min.) and too demanding (heart rate: 128 b/min; physiological cost index: 15 b/m) to be practical. Subjects supported about 1/3 of their weight with their arms during stance and about 2/3 during swing. Our results suggest that the braces reduced the effort needed and that the low speeds were due to both a lack of power at push-off and the time needed to stabilize the hip and trunk. The high heart rates arose from excessive contraction of the arm and trunk muscles for balance and propulsion.
Considerable demand exists for a device to facilitate hands-free, stable stance in individuals with neurological disorders such as spinal cord injury (SCI) and stroke. In this regard, applying functional electrical stimulation (FES) to muscles of the lower limbs in closed loop has shown promise. In particular, it has been suggested that a PID control strategy could offer functional benefits to stability by mimicking the neurological control strategy employed in able-bodied stance. In this proof of concept study, we tested this assertion by examining the potential of a PID control strategy with gravity compensation to effectively maintain balance during quiet stance by regulating FES-induced contractions of the ankle plantarflexors and dorsiflexors in able-bodied individuals. A novel Inverted Pendulum Standing Apparatus (IPSA) was employed to simulate quiet stance whilst minimizing the voluntary control of able-bodied subjects. Quiet and perturbed standing trials were performed in 3 able-bodied subjects. Performance metrics including those pertaining to stability during quiet stance (root mean square difference), perturbation rejection capabilities (settling time, peak deviation), and ability to transition from an offset initial position (settling time), were examined. For all 3 subjects and for all of the metrics examined, our results showed that the proposed closed-loop controlled FES system improved performance in comparison to voluntary control. These results indicate that the PID plus gravity control strategy used in this study offers meaningful benefits over voluntary control in terms of standing stability. Thus, the controller could potentially be applied to the problem of improving or restoring standing ability in some neurologic patient populations.
This research is in line with an important comment from the first amputee who tried the prosthetic hand with tactile feedback developed within the Smarthand project [1]. While trying the system with tactile feedback the patient said: "It's a feeling I have not had in a long time. When I grab something tightly I can feel it in the fingertips. It's strange since I don't have them anymore! It's amazing." We describe here the instrumentation and methods for testing the abilities of humans to discriminate sensations generated by electrical stimulation applied to the skin on the forearm. The instruments allowed testing of electrical stimulation with various properties (pulse duration, intensity, and rate). We tested the perception and pain thresholds, with the emphasis that comfortable sensations are a must. During the tests subjects were asked to locate the point on the skin that was stimulated and describe their perception of the elicited sensation. Results of first tests with small concentric electrodes suggest that non-amputees can distinguish up to seven perceptual qualities (the most common one was vibration, followed by tingling and tickling). Certain sensations had a higher occurrence rate along one axial line of the forearm than another of the forearm. In terms of spatial acuity, the subjects had more difficulties in distinguishing between the positions in the axial direction of the forearm compared with the circumferential direction. These results guided the design of the new array electrode with multiple cathodes and anodes positioned circumferential to the forearm. The results of the tests conducted with this electrode design showed high location discernment accuracy, and demonstrated the ability to memorize and later accurately recall six different electrical ?messages? created by delivering electrical stimulation onto three different electrode pads with two different stimulation parameters.
Stroke has a major impact in the total cost of healthcare in the Western world as stroke is the most common cause of long-term disability [1]. In attempts to enhance motor recovery after stroke effective treatment strategies have been developed in recent years. Appropriate evaluation of the intervention programs requires comprehensive and accurate assessment of the residual abnormal function. In the present study we compare two well-known clinical functional scoring tests developed for the assessment of hemiparetic upper limb function due to stroke and navigated transcranial magnetic stimulation (nTMS), which measures involuntary target muscle response to cortical stimulation. The aim is to investigate the equivalence of these methods and thus add objective evidence of the limb function to strengthen evidence-based practice. In addition to functional tests, four muscles of both arms were studied in twenty chronic stroke patients. Those patients without motor evoked potentials (MEP) to nTMS in the affected upper limb had significantly lower total score in Action Research Arm Test and Wolf Motor Function Test and longer performance time than those patients with MEP. Patients, in whom MEP in each of the four target muscles was elicitable, had better than average scores in clinical functional tests while patients, in whom no MEP was elicitable in any target muscle, had worse than average scores. Transcranial magnetic stimulation adds crucial information when clinical assessment based on voluntary activation by command is challenging, e.g. in patients suffering from cognitive deficits.
The cerebrovascular accident often results in motor impairment of one of the upper limbs, hence, compromising the quality of life of stroke survivors. Rehabilitation aims to restore the movement abilities of the paralyzed/paretic upper limb. An important element in rehabilitation is to apply a quantified measure of the quality of movement, in order to follow the recovery and select the most appropriate therapeutic modality. We developed a method that uses data recorded during planar movements and outputs an objective measure that relates to the smoothness, velocity and precision of the movement. This method is universal, in a sense that hand position can be recorded by any available means (e.g., robot assistant, digitizing board, motion tracking systems, etc). The method follows the Drawing Test, but generates results that show the ability of the patient to make point to point movements and track the presented trajectory. The method is based on measurements of hand position during movement along a target path in form of a 2 cm wide rectangle. The patient?s task is to move the hand along the target path as quickly as possible, with as few contacts (collisions) with the sides of the path. This paper addresses the aspects of automatic detection of parameters that quantify the quality of movement (speed, smoothness and precision). The use of this method is presented with 10 patients.
Mobile ad-hoc networks (MANET) are one of wireless networks implementation. MANET is very popular technology initially based on military purposes. Specific modifications of MANET created a possibility to implement several new wireless networks. One of them is a wireless mesh network (WMN). Over the last ten years, WMNs have gained more and more attention and are now considered as a convincing solution for providing better Internet access services for end users. WMN is an emerging technology that offers a cost-effective and scalable method to connect wireless devices. The main problem in WMNs is a routing protocol, especially because it has to enable the access to network for both mesh and conventional clients. Most of the existing ad hoc routing protocols use minimum hop-count as a metric for identifying the best packet routes. This paper presents neural network based approach to routing protocol for WMN. Neural networks are capable to analyze very complex network environments and solve routing problems on optimal (or almost optimal) way.
The paper addresses a problem of isolated vowels recognition in patients following total laryngectomy. The visual and acoustic speech modalities were separately incorporated in the machine learning algorithms. The authors used the Mel Frequency Cepstral Coefficients as acoustic descriptors of a speech signal. A lip contour was extracted from a video signal of the speaking faces using OpenCV software library. In a vowels recognition procedure the three types of classifiers were used for comparison purposes: Artificial Neural Networks, Support Vector Machines and Naive Bayes. The highest recognition rate was evaluated using Support Vector Machines. For a group of the laryngectomees having a different quality of speech the authors achieved 75% for acoustic and 40% for visual recognition performances. The authors obtained higher recognition rate than in a previous research where 10 cross-sectional areas of a vocal tract were estimated. Using presented image processing algorithm the visual features can be extracted automatically from a video signal.
Optical character recognition is an important task for converting handwritten and printed documents to digital format. In multilingual systems, a necessary process before OCR algorithm is script identification. In this paper novel methods for the script language identification and the recognition of Farsi handwritten digits are proposed. Our method for script identification is based on curvature scale space features. The proposed features are rotation and scale invariant and can be used to identify scripts with different fonts. We assumed that the bilingual scripts may have Farsi and English words and characters together; therefore the algorithm is designed to be able to recognize scripts in the connected components level. The output of the recognition is then generalized to word, line and page levels. We used cluster based weighted support vector machine for the classification and recognition of Farsi handwritten digits that is reasonably robust against rotation and scaling. The algorithm extracts the required features using principle component analysis (PCA) and linear discrimination analysis (LDA) algorithms. The extracted features are then classified using a new classification algorithm called cluster based weighted SVM (CBWSVM). The experimental results showed the promise of the algorithms.
In this paper is presented an investigation of the speech recognition classification performance. This investigation on the speech recognition classification performance is performed using two standard neural networks structures as the classifier. The utilized standard neural network types include Feed-forward Neural Network (NN) with back propagation algorithm and a Radial Basis Functions Neural Networks.
Focus of this paper is on the neural approach in small-signal modelling of GaAs HEMTs. Two modelling approaches based on artificial neural networks are discussed and compared. The first approach is completely based on artificial neural networks, while the second is a hybrid approach putting together artificial neural networks and an equivalent circuit representation of a microwave transistor. Both models consider the device gate width and therefore both are scalable. Results of modelling of three different AlGaAs/GaAs HEMTs in a wide range of operating bias conditions using the considered approaches are given. Different modelling aspects are discussed. A special attention is paid to the model development procedure and accuracy of the models.
In this paper we present a method for optimization of multi-pad electrode spatial selectivity during transcutaneous Functional Electrical Stimulation (FES) of hand. The presented method is based on measurement of individual muscle twitch responses during low frequency electrical stimulation via pads within multi-pad electrode. Twitch responses are recorded by Micro-Electro-Mechanical Systems (MEMS) accelerometers. The aim of this methodology is to substitute bulky sensors, torque sensors and goniometers, in multi-pad electrode optimization algorithm with smaller and lighter sensors; therefore making multi-pad stimulation suitable for daily use. Additionally we present method for minimizing number of MEMS accelerometers, which relies on characteristic waveforms of joint acceleration during wrist or fingers flexion/extension. These signals can be used to train Artificial Neural Network (ANN) to distinguish between different waveform classes and define correlation of each pad and activated muscle beneath. Results presented in this paper show high agreement of goniometers based classification and accelerometers based classification. As for classification with minimized number of sensors (one accelerometer) our ANN backed algorithm achieved high degree of accurate classification in intra-subject testing, but lower performance in inter-subject testing.
Feed-forward artificial neural networks (ANNs) have been applied to the diagnosis of mixed-mode electronic circuit. In order to tackle the circuit complexity and to reduce the number of test points hierarchical approach to the diagnosis generation was implemented with two levels of decision: the system level and the circuit level. For every level, using the simulation-before-test (SBT) approach, fault dictionary was created first, containing data relating the fault code and the circuit response for a given input signal. ANNs were used to model the fault dictionaries. At the topmost level, the fault dictionary was split into parts simplifying the implementation of the concept. During the learning phase, the ANNs were considered as an approximation algorithm to capture the mapping enclosed within the fault dictionary. Later on, in the diagnostic phase, the ANNs were used as an algorithm for searching the fault dictionary. A voting system was created at the topmost level in order to distinguish which ANN's output is to be accepted as the final diagnostic statement. The approach was tested on an example of an analog-to-digital converter, and only one test point was used i.e. the digital output. Full diversity of faults was considered in both digital (stuck-at and delay faults) and analog (parametric and catastrophic faults) part of the diagnosed system. Special attention was paid to the faults related to the A/D and D/A interfaces within the circuit.
Vestibular disorders can cause severe problems including nausea, inability to concentrate, and visual deficits. The CLONS project is developing a closed-loop sensory neural prosthesis to alleviate these symptoms. Conceptually, the prosthesis restores vestibular information by stimulating the semicircular canals according to measurements from inertial sensors rigidly affixed to the user. Here we present a project overview and brief update of our progress in animal models and selected human volunteers.
We present a method for the synthesis of electrical stimulation profiles for assisting of the walking in hemiplegic individuals. The stimulation profiles are synthesized by combining the joint torques estimated from the simulation that optimizes the tracking errors with a constraint of the minimal coactivation of antagonist muscles and the recruitment of the muscles for the potential user. The predicted electrical stimulation profiles were compared with the EMG recordings of the prime movers of the leg joints. The conclusion is that synthesis of controls should rely on muscle activation profiles determined through simulation, in which the level of coactivation of antagonist muscles are preset to ensure stability of the joints and smooth movements. The example presented uses data from a healthy individual (model parameters), but the methodology is directly implantable for hemiplegic individual just by replacing the model parameters, the EMG and the trajectory of the nonparetic leg.
We present the design, simulation and test results of a new AC amplifier for electrophysiological measurements based on a three op-amp instrumentation amplifier (IA). The design target was to increase the common mode rejection ratio (CMRR), thereby improving the quality of the recorded physiological signals in a noisy environment. The new amplifier actively suppresses the DC component of the differential signal and actively reduces the common mode signal in the first stage of the IA. These functions increase the dynamic range of the amplifier's first stage of the differential signal. The next step was the realization of the amplifier in a single chip technology. The design and tests of the new AC amplifier with a differential gain of 79.2 dB, a CMRR of 130 dB at 50 Hz, a high-pass cutoff frequency at 0.01 Hz and common mode reduction in the first stage of the 49.8 dB are presented in this paper.
We describe the hardware and software for the control of prehension for a dexterous transradial prosthesis. The prehension process comprises hand orientation (three degrees of freedom) and the opening of the hand in a manner that is appropriate for the shape and size of the object. The hardware consists of a standard web camera, accelerometer, ultrasound distance sensor, laser pointer and an LED illumination system. Software operating in real time estimates the shape and size of the object as well as the relative orientation of the hand with respect to the object. Based on this data, the controller generates signals that are sent to the three-dimensional (3D) wrist rotator, and drives which control fingers and thumb of the transradial prosthesis, thereby preparing the hand for palmar, lateral, or precision (2-digit or 3-digit) grasps. The choice of the grasp follows heuristics captured from healthy humans when grasping and expressed in the form of IF-THEN rules.
