A Mobile Brain/Body Imaging Platform for the Assessment and Optimization of Post-stroke Overground Robotic Gait Treatments

Overview

The study aims to identify neural biomarkers that can be used to monitor the recovery of locomotion and movement of post-stroke patients, both in natural settings and with enriched treatment using robotic device. These biomarkers could potentially be utilized in clinical practice to improve recovery outcomes. Additionally, the study seeks to investigate whether it is feasible to use these biomarkers to predict the extent of recovery that patients may achieve.

To achieve efficient mobility, our brain dynamically recruits different muscle subsets to create motion strategies that allow us to move and interact with the environment. Our ability to move is a result of interactions involving processes including motivation, adaptation to changes in the environment (e.g., to cope with alterations in the terrain), distractors (e.g., walking while talking over the phone) as well as motor planning and execution. These interactions are expressed through a complex control of gait parameters including speed, rhythmicity and sequencing of muscle co-activations, as well as postural adjustments needed to meet the challenge of external events. Correct information from all the sensory systems involved is synergically needed in active behavior and postural control. A wide range of causes such as neurological damage, traffic accidents and aging are responsible for locomotor deficits that greatly reduce the quality of life by considerably limiting everyday independence. Stroke is the second leading cause of death worldwide, with 6.7 million cases registered in 2012, and an estimated 50% of survivors suffering from permanent motor or cognitive impairments. Hemiparesis, one of the most common sequelae of stroke, is a leading cause of long-term disability, and often results in balance impairments and "foot drop", characterized by the inability to completely lift the toes off the ground during the swing phase of gait in more than 80% of stroke survivors. Stroke reduces the ability to recruit muscles and correctly perceive the body, often resulting in the learned non-use of the affected body part. The reduced capacity for information processing following stroke impairs the ability to efficiently distribute concentration to two or more tasks (dual task), with an increased difficulty with daily living activities that require performing multiple tasks simultaneously (e.g., walking and talking) . As sensorimotor processing capabilities are disrupted, patients are not able to interact freely with the environment as all their mental energy is "usurped" by the simple act of positioning one leg after the other. They do not have the "mental bandwidth" that we normally take for granted and that allows us for example to have a pleasant conversation with other people who are walking with us. In fact, due to the loss of physical and mental bandwidth under dual-task conditions, stroke survivors often experience falls, leading to a lack of confidence, anxiety and in some cases agoraphobia. They are also likely to experience decreased motivation and engagement in the recovery treatment, often spiraling towards a vicious circle detrimental to the success of any rehabilitation therapy attempted. Gait impairment is considered to be one of the most serious disabling sequelae of stroke, as one of the primary goals of patients is being able to walk and manage daily-life activities independently. Stroke induces modifications in locomotion, which is the primary means of daily management of mobility. Gait rehabilitation is thus the primary goal of post-stroke rehabilitation, especially given its close relation to cognitive impairment. Given the constant increase of people requiring rehabilitation or assistance treatment every year and the ever-increasing choice of treatments available, the assessment and development of suitable personalized strategies to address gait impairments is now an impending priority. Unfortunately, rehabilitation outcomes are often uncertain as they are associated with motivation, engagement and attention. Patients' engagement, in fact, will determine the effectiveness of the therapy to engage motor learning (i.e., recognition of the discrepancy between actual and expected outcomes during error-driven learning), which is an essential prerequisite for adaptation (i.e., modification of a movement, trial by trial, based on error feedback). In short, rehabilitation must be performed in a closed-loop fashion, where patients are actively involved in the rehabilitation therapy or even have a measure of control over it while receiving feedback and "rewards" that can promote sensorimotor integration and multisensory processing in the Central Nervous System (CNS). Rehabilitation acts on the mechanisms of brain plasticity, that is the intrinsic capacity of the brain to react as a highly dynamic system that can change the properties of its neural circuits. Reorganization of surviving CNS elements supports behavioral recovery, for example, through changes in interhemispheric lateralization, activity of association cortices linked to injured zones, and organization of cortical representational maps. Human-in-the-loop rehabilitation, by acting on the mechanisms of error-based and reward-based learning, can direct reorganization much more effectively towards increasing the representation of sensorimotor maps, partially (if not completely) lost after a Stroke. Conventional movement rehabilitation is mainly based on therapists' direct observation and mobilization of lower limbs, followed by assisted gait over ground, either passive or active, using context-specific motor tasks and related feedback. In "neurophysiological" techniques, the physiotherapist, based on neurophysiological knowledge of gait principles, acts as "problem solver" and "decision maker", with the patient acting as a passive recipient. In "motor learning" techniques on the other hand active patient involvement is required together with neuropsychological evaluation. To enhance patients' engagement, standardization and therapy effectiveness, enriched motor learning neurophysiological treatments that include robotic assistive devices (exoskeletons) may be used. Over the past 20 years, an overwhelming number of combinations of enriched and conventional treatments has been proposed to increase patients' motivation with interactive biofeedback or by giving patients a measure of control over the assisting robotic devices. Despite an intense debate on the merits of each solution, however, there is still insufficient evidence to clearly indicate which approach is most effective (and most importantly, why) in promoting gait recovery after stroke. In fact, we still lack the knowledge needed to apply such technology in the best possible way in relation to the individual patient's condition, especially in the case of gait rehabilitation. In the last few years Mobile Brain/Body Imaging (MoBI) has gained momentum among the scientific community as an emerging paradigm to jointly study brain and behavior and especially locomotion, also outside laboratory settings. MoBI can be used to provide the clinician with useful information to assess the rehabilitation progress by "encoding" the neural correlates of gait by means of brain-muscle connectivity assessment during movement rather than before/after the task as in current practice. Real-time MoBI may also be used to decode patient's intention of movement based on which to perform stimulus delivery (e.g., through FES, exoskeletons etc.), thus increasing engagement and promoting plastic reorganization. Successful development of MoBI setups is however still a technological challenge. Neuroimaging techniques are either not portable (e.g., Magnetic Resonance, Magnetoencephalography), or lack the temporal resolution necessary to capture near-instantaneous intra-stride modulations of neural activity during locomotion (e.g., functional near-Infrared Spectroscopy - fNIRS) because of the variation speed of the physiological marker they use (e.g., blood oxygen level for fNIRS). Electroencephalography (EEG) is the only technique that is portable enough, non-invasive and with the temporal resolution necessary to detect even such modulations of brain activity during walking. However, the EEG is highly sensitive to movement artifacts and requires solving key technological challenges and developing complex and often discouraging analysis pipelines. In any case, MoBI with EEG, though challenging, is not impossible: to date, several authors showed gait-phase-related intra-stride patterns of activation and deactivation e.g., also demonstrating a directional link between sensorimotor cortical areas and leg muscles during stereotyped gait, but studies of gait rehabilitation with MoBI neuromuscular assessment in natural ecological conditions are still critically missing. Also, the following technical/scientific question remains: "how can we distinguish movement-related neural activity from movement and mechanical artifacts? Atalante X: a disruptive wearable powered cobot. It is a collaborative exoskeleton that enables patients with severe gait impairment, including those with upper extremity dysfunction or cognition challenges to stand up and walk hands-free; it is the only cobot equipped with 12 engines at hip, knee and ankle level that allows human being to transform the intention of movement into actual movement. The project aims to optimize ATALANTE X cutting edge hi-tech solutions personalizing rehabilitation gait treatment of stroke patients using: 1. Characterization of patient's Cognitive Flexibility using Dessintey system. Dessintey is a unique technology dedicated to motor planning and central control of movement based on visuomotor simulation training that combines Action Observation-Motor Imaging-Mirror Therapy approaches. In fact, every time a movement is observed, the brain simulates instantly and without effort this same movement; in this way it is possible to increase imagination and body perception. 2. Monitoring and evaluation of Human-Cobot interaction using disruptive dynamic EEG-EMG neural biomarkers using Mobile Brain/Body Imaging (MoBI)