Rehabilitation devices based on soft robotics and bio-mechatronic sensors

Funded by University of Rijeka

Stroke is one of the world’s increasing problems, often leaving a person permanently disabled and unable to perform activities of daily living. Robotic devices are increasingly being imposed as a solution for timely and optimal rehabilitation of such patients, which, along with advanced control and measuring technologies, enable adaptive therapy and monitoring of parameters in real-time. Devices that are safe and comfortable to wear can be developed by combining the benefits of traditional and soft robotics with appropriate sensor technologies. Those systems can also enable the measurement of muscle activation as an input for adapting the amount of the generated force by an actuator. Such actuators should assure the necessary force in the rehabilitation process so the patients can perform specific movements. The proposed research represents an important segment of this chain, and contains two important directions. The first direction refers to the research and analysis of materials for developing soft robots using numerical methods based on the finite element method. Such research should result in the selection of materials with suitable mechanical properties in order to achieve the necessary functionality and durability of the component. Another important direction is the continuation of the development of a module for detecting the activation of muscle fibers of the human forearm, using machine learning and artificial intelligence methods. Namely, within the framework of the UNIRI-INOVA project whose outcomes were highly rated, significant results were accomplished and the necessary module for electromyographic (EMG) sensor signal processing was developed to achieve a high correlation with the dynamometer signal. One of the aims of this project proposal, is an extension of the functionality that allows real-time modeling and prediction of imposed force, in a very short period of time using only the EMG signal. Such a module, which would be tested experimentally among the members of the research team, should in future be a key component in the bio-mechanical interaction between a person and a rehabilitation robot. For such comprehensive research, the collaborators on the project have already applied for a competitive EU funded project.

Project results: Article “Reducing Hand Kinematics by Introducing Grasp-Oriented Intra-Finger Dependencies” published in Robotics (Bazina et al., 2024)

References

2024

  1. Robotics
    rehab1.jpg
    Reducing Hand Kinematics by Introducing Grasp-Oriented Intra-Finger Dependencies
    Tomislav Bazina, Goran Mauša, Saša Zelenika, and Ervin Kamenar
    Robotics, 2024