Search

BACKGROUND: Biomechanical gait analyses are typically performed in laboratory settings, and are associated with limitations due to space, marker placement, and tasks that are not representative of the real-world usage of lower limb prostheses. Therefore, the purpose of this study was to investigate the possibility of accurately measuring gait parameters using embedded sensors in a microprocessor-controlled knee joint.METHODS: Ten participants were recruited for this study and equipped with a Genium X3 prosthetic knee joint. They performed level walking, stair/ramp descent, and ascent. During these tasks, kinematics and kinetics (sagittal knee and thigh segment angle, and knee moment) were recorded using an optical motion capture system and force plates (gold standard), as well as the prosthesis-embedded sensors. Root mean square errors, relative errors, correlation coefficients, and discrete outcome variables of clinical relevance were calculated and compared between the gold standard and the embedded sensors.FINDINGS: The average root mean square errors were found to be 0.6°, 5.3°, and 0.08 Nm/kg, for the knee angle, thigh angle, and knee moment, respectively. The average relative errors were 0.75% for the knee angle, 11.67% for the thigh angle, and 9.66%, for the knee moment. The discrete outcome variables showed small but significant differences between the two measurement systems for a number of tasks (higher differences only at the thigh).INTERPRETATION: The findings highlight the potential of prosthesis-embedded sensors to accurately measure gait parameters across a wide range of tasks. This paves the way for assessing prosthesis performance in realistic environments outside the lab.

A semi-autonomous prosthesis control based on computer vision can be used to improve performance while decreasing the cognitive burden, especially when using advanced systems with multiple functions. However, a drawback of this approach is that it relies on the complex processing of a significant amount of data (e.g., a point cloud provided by a depth sensor), which can be a challenge when deploying such a system onto an embedded prosthesis controller. In the present study, therefore, we propose a novel method to reconstruct the shape of the target object using minimal data. Specifically, four concurrent laser scanner lines provide partial contours of the object cross-section. Simple geometry is then used to reconstruct the dimensions and orientation of spherical, cylindrical and cuboid objects. The prototype system was implemented using depth sensor to simulate the scan lines and vibrotactile feedback to aid the user during aiming of the laser towards the target object. The prototype was tested on ten able-bodied volunteers who used the semi-autonomous prosthesis to grasp a set of ten objects of different shape, size and orientation. The novel prototype was compared against the benchmark system, which used the full depth data. The results showed that novel system could be used to successfully handle all the objects, and that the performance improved with training, although it was still somewhat worse compared to the benchmark. The present study is therefore an important step towards building a compact system for embedded depth sensing specialized for prosthesis grasping.

Additional file 1. Links and descriptions to a video demonstrating the system and to the project website.

Proceedings of the 8th Graz Brain Computer Interface Conference 2019, Bridging Science and Application

Mind controlled bionic limbs promise to replace mechanical function of lost biological extremities and restore amputees’ motor capacity. State of the art approaches use machine learning for establishing a mapping function between electromyography (EMG) and joint kinematics. However, current approaches require frequent recalibration with lack of robustness, thus providing control paradigms that are sensitive to external conditions. This paper presents an alternative method based on the authors’ recent findings. That is, a biomimetic decoder comprising a computational model that explicitly synthesizes the dynamics of the musculoskeletal system as controlled by EMG-derived neural activation signals.

© 2015 Gonzalez-Vargas et al., Modern assistive devices are very sophisticated systems with multiple degrees of freedom. However, an effective and user-friendly control of these systems is still an open problem since conventional human-machine interfaces (HMI) cannot easily accommodate the system’s complexity. In HMIs, the user is responsible for generating unique patterns of command signals directly triggering the device functions. This approach can be difficult to implement when there are many functions (necessitating many command patterns) and/or the user has a considerable impairment (limited number of available signal sources). In this study, we propose a novel concept for a general-purpose HMI where the controller and the user communicate bidirectionally to select the desired function. The system first presents possible choices to the user via electro-tactile stimulation; the user then acknowledges the desired choice by generating a single command signal. Therefore, the proposed approach simplifies the user communication interface (one signal to generate), decoding (one signal to recognize), and allows selecting from a number of options. To demonstrate the new concept the method was used in one particular application, namely, to implement the control of all the relevant functions in a state of the art commercial prosthetic hand without using any myoelectric channels. We performed experiments in healthy subjects and with one amputee to test the feasibility of the novel approach. The results showed that the performance of the novel HMI concept was comparable or, for some outcome measures, better than the classic myoelectric interfaces. The presented approach has a general applicability and the obtained results point out that it could be used to operate various assistive systems (e.g., prosthesis vs. wheelchair), or it could be integrated into other control schemes (e.g., myoelectric control, brain-machine interfaces) in order to improve the usability of existing low-bandwidth HMIs., This work was supported by the Japanese Society for Promotion of Science (JSPS) Grant for Researcher Dispatch Program-International Network of Advanced Medical Engineering using Bio-robotics Technology (2009-2012), European Commission under the MYOSENS (FP7-PEOPLE-2011-IAPP-286208) project and by the German Ministry for Education and Research (BMBF) via the Bernstein Focus Neurotechnology (BFNT) Gottingen under Grant 01GQ0810.