Auswahl der wissenschaftlichen Literatur zum Thema „Transhumeral prosthesis control“

Background: Upper extremity limb loss profoundly impacts a patient’s quality of life and well-being and carries a significant societal cost. Although osseointegration allows the attachment of the prosthesis directly to the bone, it is a relatively recent development as an alternative to conventional socket prostheses. The objective of this review was to identify reports on osseointegrated prosthetic embodiment for transhumeral amputations and assess the implant systems used, postoperative outcomes, and complications. Methods: A systematic review following PRISMA and AMSTAR guidelines assessed functional outcomes, implant longevity and retention, activities of daily living, and complications associated with osseointegrated prostheses in transhumeral amputees. Results: The literature search yielded 794 articles, with eight of these articles (retrospective analyses and case series) meeting the inclusion criteria. Myoelectric systems equipped with Osseointegrated Prostheses for the Rehabilitation of Amputees implants have been commonly used as transhumeral osseointegration systems. The transhumeral osseointegrated prostheses offered considerable improvements in functional outcomes, with participants demonstrating enhanced range of motion and improved performance of activities compared with traditional socket-based prostheses. One study demonstrated the advantage of an osseointegrated implant as a bidirectional gateway for signal transmission, enabling intuitive control of a bionic hand. Conclusions: Osseointegrated prostheses hold the potential to significantly improve the quality of life for individuals with transhumeral amputations. Continued research and clinical expansion are expected to lead to the realization of enhanced efficacy and safety in this technique, accompanied by cost reductions over time as a result of improved efficiencies and advancements in device design.

ABSTRACT Introduction To choose a suitable prosthesis, clients need to experience both the weight and the control of a prosthesis. A few years ago, De Hoogstraat Rehabilitation Center developed the Xperience Prosthesis for children and adults with a transradial congenital or acquired limb deficiency. Because of the positive effects, we developed a reusable test prosthesis for the transhumeral level. Xperience Prosthesis Transhumeral is an innovative test prosthesis and an essential tool in managing expectations when providing clients with a suitable upper-limb prosthesis. Xperience Prosthesis Transhumeral is a 3D-printed, reusable adjustable socket system with a passive elbow unit and the possibility to fit and experience myoelectric, static, and passive terminal devices. Conclusions Xperience Prosthesis Transhumeral is a practical and easy-to-handle tool for professionals. For clients, this tool is a unique way to experience the weight, function, and control of a prosthesis before making a final choice. Clinical Relevance Using Xperience Prosthesis Transhumeral in an expert rehabilitation center for upper-limb clients guides professionals to choose the right prosthesis with the client.

Prosthetic arms are designed to assist amputated individuals in the performance of the activities of daily life. Brain machine interfaces are currently employed to enhance the accuracy as well as number of control commands for upper limb prostheses. However, the motion prediction for prosthetic arms and the rehabilitation of amputees suffering from transhumeral amputations is limited. In this paper, functional near-infrared spectroscopy (fNIRS)-based approach for the recognition of human intention for six upper limb motions is proposed. The data were extracted from the study of fifteen healthy subjects and three transhumeral amputees for elbow extension, elbow flexion, wrist pronation, wrist supination, hand open, and hand close. The fNIRS signals were acquired from the motor cortex region of the brain by the commercial NIRSport device. The acquired data samples were filtered using finite impulse response (FIR) filter. Furthermore, signal mean, signal peak and minimum values were computed as feature set. An artificial neural network (ANN) was applied to these data samples. The results show the likelihood of classifying the six arm actions with an accuracy of 78%. The attained results have not yet been reported in any identical study. These achieved fNIRS results for intention detection are promising and suggest that they can be applied for the real-time control of the transhumeral prosthesis.

The design of artificial limbs is a research topic that has, over time, attracted considerable interest from researchers in various fields of study, such as mechanics, electronics, robotics, and neuroscience. Continuous efforts are being made to build electromechanical systems functionally equivalent to the original limbs and to develop strategies to control them appropriately according to the intentions of the user. The development of Human–Machine Interfaces (HMIs) is a key point in the development of upper limb prostheses, since the actions carried out with the upper limbs lack fixed patterns, in contrast to the more predictable nature of lower limb movements. This paper presents the development of an HMI system for the control of a transhumeral prosthesis. The HMI is based on a hybrid control strategy that uses voice commands to trigger prosthesis movements and regulates the applied grip strength when the user turns his head. A prototype prosthesis was built using 3D technology and trials were conducted to test the proposed control strategy under laboratory conditions. Numerical simulations were also performed to estimate the grip strength generated. The results obtained show that the proposed prosthesis with the dedicated HMI is a promising low-cost alternative to the current solutions. The proposed hybrid control system is capable of recognizing the user’s voice with an accuracy of up to 90%, controlling the prosthesis joints and adjusting the grip strength according to the user’s wishes.

Transhumeral amputees experience considerable difficulties with controlling a multifunctional prosthesis (powered hand, wrist, and elbow) due to the lack of available muscles to provide electromyographic (EMG) signals. The residual limb motion strategy has become a popular alternative for transhumeral prosthesis control. It provides an intuitive way to estimate the motion of the prosthesis based on the residual shoulder motion, especially for target reaching tasks. Conventionally, a predictive model, typically an artificial neural network (ANN), is directly trained and relied upon to map the shoulder–elbow kinematics using the data from able-bodied subjects without extracting any prior synergistic information. However, it is essential to explicitly identify effective synergies and make them transferable across amputee users for higher accuracy and robustness. To overcome this limitation of the conventional ANN learning approach, this study explicitly combines the kinematic synergies with a recurrent neural network (RNN) to propose a synergy-space neural network for estimating forearm motions (i.e., elbow joint flexion–extension and pronation–supination angles) based on residual shoulder motions. We tested 36 training strategies for each of the 14 subjects, comparing the proposed synergy-space and conventional neural network learning approaches, and we statistically evaluated the results using Pearson’s correlation method and the analysis of variance (ANOVA) test. The offline cross-subject analysis indicates that the synergy-space neural network exhibits superior robustness to inter-individual variability, demonstrating the potential of this approach as a transferable and generalized control strategy for transhumeral prosthesis control.

The control scheme in a myoelectric prosthesis includes a pattern recognition section whose task is to decode an input signal, produce a respective actuation signal and drive the motors in the prosthesis limb towards the completion of the user’s intended gesture motion. The pattern recognition architecture works with a classifier which is typically trained and calibrated offline with a supervised learning framework. This method involves the training of classifiers which form part of the pattern recognition scheme, but also induces additional and often undesired lead time in the prosthesis design phase. In this study, a three-phase identification framework is formulated to design a control architecture capable of self-learning patterns from bio-signal inputs from electromyography (neuromuscular) and electroencephalography (brain wave) biosensors, for a transhumeral amputee case study. The results show that the designed self-learning framework can help reduce lead time in prosthesis control interface customisation, and can also be extended as an adaptive control scheme to minimise the performance degradation of the prosthesis controller.

Background: Although targeted muscle reinnervation has been shown to be effective in enhancing prosthetic control for upper limb amputees, restored hand sensations have been variable. An understanding of possible sensory feedback channels is crucial in working toward more effective closed-loop prosthetic control. Objectives: To compare sensory outcomes of different targeted sensory reinnervation approaches. Study design: Case series, cross-sectional, and retrospective. Methods: Three transhumeral amputees that had undergone different sensory reinnervation approaches were recruited. Skin pressure sensitivity thresholds and anatomic sensory mapping were performed using Semmes-Weinstein monofilaments. The clinical charts of the subjects were reviewed to compare the sensory maps performed during the earlier post-reinnervation period. Results: While the first two subjects achieved return of hand sensations on the stump skin in early follow-up, the maps showed attenuation over time. The last subject developed discrete sensations of all digits in the recipient cutaneous nerve territories away from the reinnervated muscles. Conclusions: These findings confirm that it is feasible to restore hand sensation after transhumeral targeted reinnervation, but there is a significant intersubject variability. The intrafascicular approach may be particularly effective in restoring digit sensation and deserves further exploration, as do factors affecting stability of the hand maps over time. Clinical relevance In addition to enabling intuitive motor control of myoelectric prosthesis, targeted reinnervation can also result in sensory restoration of the hand. Documentation of sensory mapping present after reinnervation may assist with exploring future techniques for sensory enhancement, with the goal of working toward closed-loop prosthetic control.

Les progrès de la mécatronique ont permis d’améliorer les prothèses du membre supérieur en augmentant le catalogue des mouvements prothétiques. Cependant, un fossé se creuse entre les capacités technologiques de la prothèse et leur méthode de contrôle. La commande myoélectrique, qui est la méthode la plus répandue, reste complexe, notamment pour les personnes amputées au niveau trans-huméral qui peuvent avoir un coude actif en plus de la main et du poignet motorisés. Une approche intéressante consiste à utiliser la mobilité du membre résiduel, présente chez la plupart des amputés trans-huméraux, pour contrôler des articulations prothétiques distales comme le coude. Les mouvements du coude sont couplés aux mouvements du membre résiduel selon un modèle de coordination épaule/coude saine. Cette thèse étudie une stratégie de commande d’un coude prothétique utilisant les mouvements du membre résiduel, mesuré par des centrales inertielles, et nos connaissances du contrôle moteur humain. Pour cela, un modèle de la coordination épaule/coude a été construit à partir d’enregistrements de gestes sains de préhension. Ce modèle, implémenté sur un prototype de prothèse, a été testé par 10 individus sains équipés du prototype afin de valider le concept, puis par 6 personnes amputées. Ces dernières ont aussi réalisé la tâche avec une commande myoélectrique conventionnelle afin de comparer les résultats. La commande couplant automatiquement les mouvements de l’épaule et du coude s’est montrée satisfaisante en termes de facilité d’utilisation et de réduction des stratégies de compensation
Progress in mechatronics has enabled the improvement of upper limb prosthetics increasing the grasps catalog. However, a gap has been growing between the prosthesis technological possibilities and the methods to control it. Indeed, common myoelectric control strategy remains complex, especially for transhumeral amputees who can have an active elbow in addition to a prosthetic wrist and hand. Since most transhumeral amputees have a mobile residual limb, an interesting approach aims at utilizing this mobility to control intermediate prosthetic joints, like the elbow, based on the shoulder/elbow coordination observed in healthy movements. This thesis investigates the possibility of controlling an active prosthetic elbow using the residual limb motion, measured with inertial measurement units, and knowledge of the human motor control. A primary focus has been targeting the reaching movement for which a model has been built using regression tools and kinematic data from several healthy individuals. The model, implemented on a prosthesis prototype, has been tested with 10 healthy participants wearing the prototype to validate the concept, and with 6 amputated individuals. These participants also performed the task with a conventional myoelectric control strategy for comparison purpose. The results show that the inter-joint coordination-based control strategy is satisfying in terms of intuitiveness and reduction of the compensatory strategies

Les contrôles myoélectriques pour les prothèses transhumérales entraînent souvent un taux élevé d’abandon en raison de leurs performances insatisfaisantes. Inspirés des progrès réalisés dans les contrôles exploitant les mouvements résiduels, nous avons affiné une approche alternative utilisant un réseau de neurones artificiels entrainé sur les mouvements naturels de bras pour prédire la configuration des articulations distales en fonction du mouvement des articulations proximales et d’information sur l’objet à saisir. Des études antérieures ont montré que cette stratégie permet aux amputés de contrôler un avatar de prothèse dans un environnement de réalité virtuelle aussi bien qu’avec leur bras valide. Cependant, le déploiement de ce contrôle dans des scénarios réels requiert des développements supplémentaires. Il est nécessaire d’intégrer une caméra montée sur la tête et des algorithmes de vision par ordinateur pour estimer en temps réel la position et l’orientation de l’objet. Dans ce contexte, les informations sur l’objet ne seraient disponibles que dans un référentiel centré sur la tête de l’utilisateur, alors que notre contrôle repose sur l’objet exprimé dans un référentiel centré sur l’épaule. Inspirés de la façon dont le cerveau exécute des transformations de coordonnées, nous avons développé et testé des solutions pour effectuer la transformation tête-épaule à partir des seules données d’orientation, disponibles en situation réelle. Pour développer ces algorithmes, nous avons constitué une base de données incluant la relation entre ces référentiels en demandant à vingt participants valides de saisir des objets dans diverses positions et orientations dans un environnement virtuel. Cette base de données comprenait les mouvements de la tête et du regard, ainsi que ceux du tronc, des épaules et des bras, capturant l’ensemble de la chaîne cinématique entre le but du mouvement et la main déplacée pour l’atteindre. Ensuite, nous avons mis en œuvre deux méthodes pour obtenir la position de la tête dans le référentiel de l’épaule. La première consiste en un réseau de neurones artificiels entraîné hors ligne sur la base de données pour prédire cette position en fonction de la taille du participant et de l’orientation de sa tête et de son épaule. La seconde méthode s’inspire des processus d’intégration multisensorielle du cerveau et déduit la position de la tête dans le référentiel de l’épaule en comparant les données relatives à la main prothétique obtenues dans le référentiel de l’épaule par cinématique directe et simultanément dans le référentiel de la tête par la vision par ordinateur. Inspirés par les mécanismes neuronaux du codage de l’espace péripersonnel, nous avons adapté en ligne les poids d’un réseau de neurones pour coder cette différence dans une carte spatiale. Les résultats expérimentaux sur douze participants valides en réalité virtuelle ont démontré des erreurs persistantes avec la première méthode, qui n’a pas réussi à prendre en compte avec suffisamment de précision la spécificité de la morphologie de l’utilisateur, entraînant ainsi un contrôle inefficace de la prothèse. En revanche, la seconde méthode a réussi à coder efficacement la transition de la tête à l’épaule associée à différentes cibles dans l’espace. L’efficacité de la seconde méthode a également été testée sur six amputés en réalité virtuelle, et une preuve de concept a été réalisée pour évaluer sa faisabilité en conditions réelles. Cette démonstration a été réalisée en contrôlant la plateforme robotique REACHY 2 en vision égocentrée, avec des marqueurs ArUco et un algorithme de vision artificielle pour détecter les objets à saisir et la main robotique. Les résultats suggèrent que, malgré les difficultés rencontrées dans la détection des objets, notre carte spatiale fonctionne efficacement dans des scénarios réels. Cette méthode pourrait également gérer des scénarios complexes, impliquant des déplacements de caméra ou des environnements perturbés
Myoelectric controls for transhumeral prostheses often lead to high rates of device abandonment due to their unsatisfactory performance. Grounded on advances in movement-based prosthesis control, we refined an alternative approach using an artificial neural network trained on natural arm movements to predict the configuration of distal joints based on proximal joint motion and movement goals. Previous studies have shown that this control strategy enabled individuals with transhumeral limb loss to control a prosthesis avatar in a virtual reality environment as well as with their valid arm. Yet, deploying this control system in real-world requires further development. A head-mounted camera and computer vision algorithms need to be integrated into the system for real-time object pose estimation. In this setup, object information might only be available in a head-centered reference frame, while our control relies on the object expressed in a shoulder reference frame. Taking inspiration from how the brain executes coordinate transformations, we developed and tested solutions to perform the required head-to-shoulder transformation from orientation-only data, possibly available in real-life settings. To develop these algorithms, we gathered a dataset reflecting the relationship between these reference frames by involving twenty intact-limbs participants in picking and placing objects in various positions and orientations in a virtual environment. This dataset included head and gaze motion, along with movements of the trunk, shoulders, and arm joints, capturing the entire kinematic chain between the movement goal and the hand moved to reach it. Following data collection, we implemented two methods to transform target information from the head to the shoulder reference frame. The first is an artificial neural network trained offline on the dataset to predict the head position in the shoulder referential given ongoing shoulder and head orientations and the participant height. The second method draws inspiration from multisensory integration in the brain. It derives the head position in the shoulder referential by comparing data about the prosthetic hand obtained in the shoulder referential through forward kinematics and simultaneously in the head referential through computer vision. Inspired by brain’s mechanisms for peripersonal space coding, we encoded this difference in a spatial map by adapting the weights of a single-layer network of spatially tuned neurons online. Experimental results on twelve intact-limbs participants controlling a prosthesis avatar in virtual reality demonstrated persistent errors with the first method, which failed to adequately account for the specificity of the user’s morphology, resulting in significant prediction errors and ineffective prosthesis control. In contrast, the second method elicited much better results and effectively encoded the transition from the head to the shoulder associated with different targets in space. Despite requiring an adaptation period, subsequent performances on already explored targets were comparable to the ideal scenario. The effectiveness of the second method was also tested on six participants with transhumeral limb loss in virtual reality, and a physical proof of concept was implemented on a teleoperated robotic platform with simple computer vision to assess feasibility in real-life settings. One intact-limbs participant controlled the robotic platform REACHY 2 to grasp cylinders on a board. ArUco markers on the robot’s end effector and cylinders coupled with a gaze-guided computer vision algorithm enabled precise object pose estimation. The results of this proof of concept suggest that despite challenges in object detection, our bio-inspired spatial map effectively operates in real-world scenarios. This method also shows promise for handling complex scenarios involving errors in position and orientation, such as moving a camera or operating in perturbed environments

Development in the field of bio-mechatronics has provided diverse ways to mimic and improve the function of human limbs. Without an elbow joint, the hand remains stiff because all the muscles tension passes through this joint. Advanced myoelectric prosthetic devices are limited due to the lack of appropriate signal sources on residual amputee muscles and insufficient real-time control. Neural-machine interfaces (NMI) are representing a recent approach to develop effective applications. In this research study, an NMI is designed that presents real-time signal processing for command generation. The human brain hemodynamic responses are, therefore, translated into control commands for people suffering from transhumeral amputation. A novel and first of its kind scheme is proposed which utilizes functional near-infrared spectroscopy (fNIRS) to generate the control commands for a three-degree-of-freedom (DOF) prosthetic arm. The time window for fNIRS signals was set to 1 second. The average accuracy was found to be 82% which is a state-of-the-art result for such a technique. The accuracy ranged from 65 to 85% subject-wise. The data were trained and tested on both artificial neural network (ANN) and linear discriminant analysis (LDA). Eight out of 10 motions were correctly predicted in real time by both classifiers.