Duchenne Muscular Dystrophy (DMD) patients suffer from a severe form of progressive muscular weakness. Consequently, as the disease progresses these patients become more and more dependent on assistive devices for their daily activities. For instance lifting their arms against gravity already becomes challenging. So far, a considerable amount of effort has been put in the development of assistive devices. However, compensating the weight of the hand remains challenging as the required balancing torques are dependent on the position of the hand and the orientation of the forearm. In this research a solution is proposed to passively compensate the weight of the hand by making use of a simplified torque profile (a constant torque). The effectiveness of this profile was assessed experimentally in a group of healthy subjects. For this the wrist muscle activity required to lift the hand against gravity was measured through surface electromyography, for several types of compensation. From this experiment the conclusion can be drawn that a constant torque profile performs similar to the theoretically ideal type of balancing, showing a similar reduction in the anti-gravity muscle activity. Based on these findings, a mechanism has been developed capable of providing an adjustable constant force, by making use of a combination of positive and negative stiffness springs, for implementation in a wrist support for DMD patients.

Neuromuscular diseases often result in elevated passive joint impedance (PJI), impacting the daily lives of affected individuals. To be able to apply assistive devices that compensate for PJI, it is essential to identify the PJI correctly. This study aims to identify and model the PJI of the elbow joint, to improve the control of force-based controlled active-assistance exoskeleton devices by distinguishing the voluntary from the passive forces. We used an elbow device to measure the PJI in twelve healthy males with the shoulder abducted at a 90˚ angle. The study investigates both static and dynamic conditions, encompassing various contraction velocities (≤ 0.20 rad/s) and shoulder flexion positions (0˚ and ±45˚). The analysis estimates the hysteresis, equilibrium position, and elastic property. Subsequently, based on the average elbow PJI, we developed a regression model, with the statistical analysis revealing no significant differences between the conditions, except for the observed hysteresis under varying velocities. Based on the tested conditions, the findings indicate that a single low-velocity dynamic measurement can serve as the basis for deriving a general elbow PJI regression model.

Severe muscle weakness is a symptom appearing in certain neuromuscular diseases (NMDs), such as Duchenne Muscular Dystrophy (DMD), affecting people's daily lives by reducing functionality, decreasing independence, and reducing the ability to perform essential daily activities. This patient group might benefit from using active-assistive devices by having the potential to provide precise support torque counterbalancing the passive forces acting on the arm, the movement intention of the user, and external forces exerted by lifted objects. However, the determination of support to counteract the weight of lifted objects is an ongoing challenge. This research aims to improve the understanding of external forces by using data classification algorithms to distinguish between different lifted weights in a human experiment. Fourteen healthy individuals participated in this experiment, lifting weights ranging from 0 - 1000 grams while an active-assistive device compensated for the passive torques acting on the arm. Data was collected using various sensors: a force sensor, an Inertial Measurement Unit (IMU), a joint encoder, and surface Electromyography (sEMG) electrodes. Subsequently, this data was processed and fed into a K Nearest Neighbour (KNN) classifier and a Support Vector Machine (SVM) classifier to determine the lifted weights during human elbow flexion and extension. The classifier showing the highest performance achieved an accuracy of 39.70% on the test dataset, indicating several misclassifications. However, a recall percentage of 76.95% for the 1000-gram class within the multi-class classification demonstrates the capability to distinguish larger weights. While demonstrating potential in weight discrimination, especially for larger weights, improvements in the compensation strategy, arm support alignment, and experimental design are crucial. Future research on the impact of picking and placing objects, the influence of muscle weakness, and the application of alternative data classification algorithms are essential to further enhance understanding of the interaction with objects and result in more accurate predictions.

DMD (Duchenne muscular dystrophy) is a genetic disorder characterized by progressive muscle weakness, leading to the eventual loss of muscle function. After losing lower extremity function, DMD patients lose the ability to use their arms. To assist individuals with DMD, an upper extremity exoskeleton is being developed to provide support. The device is required to provide proper weight compensation for the weight of both the user and the support system. Previous literature review has highlighted three strategies for achieving compensation: Model, Calibration dynamic and Calibration static. However, the review lacks a conclusive understanding of the differences between these strategies. Thus, this study aims to validate a weight compensation model by comparing the joint torque with two different measurements: Calibration static and Calibration dynamic on non-disabled participants with a one DOF (Degree of freedom) elbow support system. The weight compensation model was designed to accurately account for the weight of the user's arms and the exoskeleton device itself. It incorporates multiple inputs, including shoulder flexion, shoulder abduction, elbow joint angle, arm mass, and arm center of mass to calculate the required compensation torque for the motor located at the elbow joint. The model was validated by a dead weight experiment. The weight compensation model was validated by comparing the joint torque measurements from Calibration dynamic and Calibration static results. Afterwards, experiments were performed on 12 male non-disabled participants. The weight compensation model results does not align well with the measurements from non-disabled participants. Analysis suggests that joint impedance caused these discrepancies. However, even after accounting for joint impedance, the weight compensation model still exhibits a tendency to overestimate the required compensation torque. A fitted model was used to decrease the product value of mass and center of mass to decrease overestimation. Furthermore, the comparative analysis indicates that dynamic and static measurements yielded similar mean values for joint torque. While the weight compensation model demonstrates accuracy in a dead weight experiment under the assumption of an accurate estimation of its mass and center of mass, its performance is sub-optimal during the human experiment. This is attributed to the absence of joint impedance consideration and the overestimation of the forearm plus hand mass and center of mass for the user. Comparison between dynamic and static measurements the mean difference between dynamic and static measurements obtained from non-disabled participants indicates no substantial disparity in terms of joint torque. Future plans for this research involve expanding the current one DOF configuration to a four DOF configuration and incorporating an inertial measurement unit (IMU) for more accurate angle measurements. Additionally, different compensation strategies will be compared in terms of task performance using metrics such as external interaction force or sEMG (surface electromyography). Lastly, the overestimation of forearm plus hand mass and center of mass will be further investigated.