IMU-based monitoring system for a stroke rehabilitation knee-exoskeleton

One of the fields in biomedical engineering that greatly improved in the last decades is the one related to the interaction between robotic systems and the human body in assistance and rehabilitation, as they represent a valuable solution for motor-impaired people. These technologies are also known as mobility assistive technologies. Exoskeletons represent one of the most recent solution to provide assistance in performing motion. However in most cases, they are still bulky, they need a high number of instruments and tools to work properly, both from the hardware and software perspective. This means that their use is often limited to research laboratories or rehabilitation centres. These systems are intended to allow people with motor disabilities to recover the natural motor behaviour. Therefore, it is important to allow their use also into contexts of everyday life, such as office, school, and home. The present work is part of the Enhance Project, which aims to develop a knee exoskeleton for the rehabilitation of stroke patients that can also be used at home. The main goal of this thesis consists in defining the viability of a monitoring system, composed of three inertial measurement units (IMUs) and an encoder. This setup is necessary to evaluate kinematic quantities such as accelerations and joint angles, that are used to efficiently control the robotic device. In the first place, the state of the art on lower-limb rehabilitative exoskeletons, or other robotic devices, for stroke patients that can be used at home, or in a more generic community context, is assessed. Considering the monitoring system of the exoskeletons, IMUs are widely used for these devices, as well. One healthy subject is chosen to wear the entire system (devices plus IMUs), to assess the reliability of the quantities obtained by the inertial sensors. Before that, to assess that the system is correctly working, we built a test bench to perform simple movements with the IMUs to check their accuracy. The validation is firstly carried out with the use of a goniometer, then we use an encoder as a more valuable tool of validation. Indeed, the optical sensor is embedded in the motor that controls the knee flexion-extension, while the sensors are attached to the robotic device itself. Also, as a further validation system, another group of IMUs (XSens DOT), already available in the market, is used. Finally, some trials on a human subject are then carried out to see if the entire exoskeleton system is suitable to be used. The results show that the embedded monitoring system can be very accurate, as in the cases of trials with the validation through encoder data, it follows the optical sensor with high precision.