About a year ago (to be more precise, a year and 3 months ago), I applied for a grant at the UEFISCDI research agency. The initial plan was to have the results by May 2017 and start working on the projects sometime during the summer. Long story short, we had quite some delays, receiving news about the ranking only a few weeks ago.
Last week, we got the final answer, and the news were good: The government had allocated the budget for the projects for this year and our proposal was among the funded ones! The wait was long, but at least it seems that it was worth it.
So what is the project all about? I will try to document it through various posts in this page, but here is a brief overview of the idea of the project.
An increasing amount of people, especially elderly ones, face mobility problems often due to a stroke or some kind of accident. To increase their quality of life and reduce their dependence on others, physical rehabilitation is necessary. The aim of the rehabilitation is to enable the people to re-learn how to use their limbs and strengthen the affected muscles. The traditional way of performing rehabilitation is by helping the patients pefrorm repetitive motions. A physiotherapist is monitoring their effort and progress and adjusts the task and the motion respectively. However, due to the fact that different patients have different mobility problems, the pre-defined repetitive motions do not constitute an optimal rehabilitation scheme for everyone. In other words, the rehabilitation is not always personalised.
Our proposal is about designing a __B__iomecanically E__nabled robo__T__ic controll__ER for __RE__storing __H__uman __AB__ility (BETER REHAB). The idea is to use a robotic arm to improve the rehabilitation process of a patient. The robotic arm will be attached on the arm of the patient and will firstly detect the intention of motion of the patient. Based on this intention, it will calculate a trajectory of motion and will apply the appropriate force along that trajectory. This way, the patient will be free to perform any type of motion, not just the predefined ones, and receive the appropriate assistance from the robotic arm.
To be able to do this, the robotic controller needs to be aware of the biomechanics of the arm. Therefore a musculoskeletal model of the upper arm will be included in the feedback loop of the controller of the robot. To better estimate and understand the intention of motion of the patient, we will simultaneously measure the muscle activation of the upper arm and the current position of the arm segments. Finally, the idea is to apply enforced learning techniques to allow calibration of the system for each patient individually.
I am getting very excited about starting my work on this project. It seems like a lot of work that needs to be accomplished within two years, but at least I will have great people to support me with this. Lucian Bușoniu will be the supervisor of this project and will bring his expertise in control and learning. Levente Tamas, even though not officially attached to the project, has already helped with the vision aspects of project. And of course, we will receive a lot of valuable feedback from our industrial partners. Polaris, a rehabilitation clinic in Cluj, will not only provide clinical feedback, but also help us recruit patients for trial runs. And finally MIRA Rehab will bring their expertise on rehabilitation through gaming and possible ways on how to integrate our platform with theirs.
To be continued….!