A robot-aided evaluation of the biomechanical impedance of the wrist as an assessment tool of spasticity

According to the World Health Organization, approximately 15% of the world's population has some form of disability. Therefore, objective assessment and subsequent tailored rehabilitation treatments become crucial in decreasing patients' level of disability. A promising solution to achieve this goal is to bring advanced technologies, such as robotic devices, into clinical settings and apply them as routine assessment and rehabilitation tools. However, standardized and quantitative measures are still lacking despite the well-known advantage of using robotic devices. The most accepted and used assessment methods in clinical settings are manual scales. Indeed, manual assessments are preferred to any other measurements since they require very little time to set up and often do not involve equipment. Nevertheless, such scales have several limitations, such as low reliability and sensitivity. From an extensive study designed to identify the use and quantify the validity of robotic assessment in the clinical setting, we realized that one of the missing parts was the lack of objectivity in assessing the spasticity condition. Spasticity affects many patients with neurological disorders. It can drastically impair their voluntary motor control, limiting their independence and making difficult their activities of daily living. Therefore, a quantitative and reliable measurement of spasticity can play a crucial role in defining the best rehabilitation protocol and subsequent monitoring of patient progress and therapeutic efficacy. Among other things, spasticity involves exaggerated stretch reflex responses and abnormal increases in muscle tone. In this context, we implemented a protocol on a robotic device that was able to record quantitative measures and be easily transferred to the clinical setting. In particular, we developed a novel protocol to assess the mechanical impedance of the wrist in response to precise position-controlled perturbations provided by the robot. The primary objectives of the protocol were to standardize the process of assessing the wrist's biomechanical properties and minimize measurement variability. Because variability in measurements can depend on several factors, such as grip force and the occurrence of muscle fatigue, we added to the implemented protocol the ability to consider both of these variables during the assessment session. Furthermore, we cared to propose such a protocol that was also feasible in the clinical setting. In the first part of the dissertation, we tested the implemented protocol in a healthy population. We verified its reliability and built a large reference dataset of the biomechanical properties of the wrist under different experimental conditions to allow subsequent comparisons with patients. Subsequently, in the second part, we collected data on a clinical population, and despite the difficulties in recruiting patients during the pandemic, the clinical feasibility was very promising. Among others, one of the advantages of this protocol lies in its duration: it lasts about 5 minutes, including recording the grip force exerted during the assessment session and detecting the onset of fatigue. Furthermore, special attention was given to the WristBot throughout the dissertation. The robot, derived from a research prototype, has been entirely re-engineered to be nearly ready for transfer into the clinical setting.