Quantitative Assessment of Training Effects Using EKSOGT Exoskeleton in Quantitative Assessment of Training Effects Using EKSOGT Exoskeleton in Parkinson Disease Patients (Ekso_PD)

Quantitative Assessment of Training Effects Using EKSOGT Exoskeleton in Quantitative Assessment of Training Effects Using EKSOGT Exoskeleton in Parkinson Disease Patients

Quantitative Assessment of Training Effects Using a Wearable Exoskeleton in Parkinson Disease Patients

Fresco Parkinson Center Villa Margherita, Vicenza, Italy, Fresco Institute for Parkinson's & Movement Disorders, NYU Langone

The ability to walk independently is a primary goal when rehabilitating an individual with Parkinson Disease (PD). Indeed, PD patients display a flexed posture that coupled with an excessive joint stiffness lead to a poor walking mechanics that increase their risk of falls. Although studies have already shown the many benefits of robotic-assisted gait training in PD patients, research focusing on optimal rehabilitation methods has been directed towards powered lower-limb exoskeleton. Combining the advantages delivered from the grounded devices with the ability to train in a real-world environment, these systems provide a greater level of subject participation and increase subject's functional abilities while the wearable robotic system guarantees less support. The purpose of the present work is to evaluate the effects of an Over-ground Wearable Exoskeleton Training (OWET) on gait impairments in comparison with a multidisciplinary intensive rehabilitation treatment. As gait is a complex task that involves both central (CNS) and peripheral nervous systems (PNS), targeted rehabilitation must restore not only gait mechanics (ST parameters) but also physiological gait pattern (joint kinematics and dynamics). To this aim the impact of OWET on both CNS and PNS will be evaluated. Thus, a quantitative assessment of an individual's gait and neuromuscular function to robustly evaluate recovery of altered sensorimotor function at both the PNS and CNS is proposed. To this aim, comprehensive GA (spatiotemporal (ST) parameter, joint kinematics, joint stiffness) and electromyography (EMG) will be combined to determine PNS improvements, and fMRI with EEG will be used to assess CNS improvements.

Full Title: QUANTITATIVE ASSESSMENT OF TRAINING EFFECTS USING A WEARABLE EXOSKELETON IN PARKINSON DISEASE PATIENTS RESEARCH PLAN Specific Aims The ability to walk independently is a primary goal when rehabilitating an individual with Parkinson Disease (PD). Indeed, PD patients display a flexed posture that coupled with an excessive joint stiffness lead to a poor walking mechanics that increase their risk of falls. Although studies have already shown the many benefits of robotic-assisted gait training in PD patients, research focusing on optimal rehabilitation methods has been directed towards powered lower-limb exoskeleton. Combining the advantages delivered from the grounded devices with the ability to train in a real-world environment, these systems provide a greater level of subject participation and increase subject's functional abilities while the wearable robotic system guarantees less support. The purpose of the proposed work is to evaluate the effects of an Over-ground Wearable Exoskeleton Training (OWET) on gait impairments in comparison with a multidisciplinary intensive rehabilitation treatment. As gait is a complex task that involves both central (CNS) and peripheral nervous systems (PNS), targeted rehabilitation must restore not only gait mechanics (ST parameters) but also physiological gait pattern (joint kinematics and dynamics). To this aim the impact of OWET on both CNS and PNS will be evaluated. Human movement analysis quantitatively assesses the neuromuscular and biomechanical features of movement. Recent literature has highlighted the benefit of coupling gait analysis (GA) and neuromusculoskeletal modeling (NMSM) for treatment planning and supplementing this approach with robotic rehabilitation. Another stalwart of PD research has been electroencephalography (EEG), which is widely used to evaluate executive dysfunction while functional magnetic resonance imaging (fMRI) can detect cortical changes in motor activations during motor tasks. Thus, a quantitative assessment of an individual's gait and neuromuscular function to robustly evaluate recovery of altered sensorimotor function at both the PNS and CNS is proposed. To this aim, comprehensive GA (spatiotemporal (ST) parameter, joint kinematics, joint stiffness) and electromyography (EMG) will be combined to determine PNS improvements, and fMRI with EEG will be used to assess CNS improvements. As health care professionals and researchers need objective, reliable, and valid tools to plan subject-specific interventions, quantify therapeutic outcomes, and monitor change over time, the proposed study includes estimation of neutrally-informed muscle forces and joint stiffness, which is expected to provide sensitive determinants of PD movement control that could be crucial to inform treatment planning/assessment. Preliminary data are available and showed feasibility of the proposed measurement set up. Background OWET: Although studies have already shown the many benefits of robotic-assisted gait training in PD patients (i.e. body weight supported treadmill training) as improving gait efficiency modifying spatiotemporal (ST) parameters, these strategies create an environment where the patient has less control over the gait initiation and lacks in variability of visuospatial flow. Therefore, research focusing on optimal rehabilitation methods has been directed towards powered lower-limb exoskeleton, as in post-stroke rehabilitation, where the effect of such a treatment dramatically enhanced potential for patient-specific rehabilitation, showing improvement in ST parameters. Combining the advantages delivered from the grounded robotic devices with the ability to train the patient in a real-world environment, these systems provide a greater level of subject participation for maintaining trunk and balance control, as well as navigating their path over different surfaces and increase subject's functional abilities while the wearable robotic system guarantees less support. Furthermore, the stability the exoskeleton addresses to the patient, allows a hands-free walking trial (with no clutches) which represents an integral part for a physiological locomotion restoration. As gait involves both CNS and PNS, targeted rehabilitation must restore not only mechanics (speed, stride time and length) but also physiological gait pattern. This requires improvements at the level of both balance and lower limb joint motion. In this direction, wearable lower-limb powered exoskeletons promote functional training in a realistic walking-environment combined with a greater patient's engagement than in grounded devices. Human movement analysis quantitatively assesses the neuromuscular and biomechanical features of movement. Recent literature has highlighted the benefit of coupling GA and NMSM for treatment planning and supplementing this approach with robotic rehabilitation, however there is no study investigating gait effects from an OWET in those with PD, and no assessment that uses comprehensive GA and NMSM to reveal mechanistic changes as a result of therapy. Neurophysiology of PD: A stalwart of PD research has been EEG, which is widely used to evaluate executive dysfunction while fMRI can detect cortical changes in motor activations during motor tasks. The protocol to use GA in combination with fMRI has already been adopted by the investigators in order to display the impacts of the rehabilitation process on the reorganization of the neural network, describing and quantifying the neural activity and the recovery after the treatment. Motion analysis in PD: Gait in people with PD has been thoroughly studied with 3D GA systems in recent years, documenting a typical hypokinetic gait reduction of the stride-length with asymmetry between the strides, an increment of the cadence, the stance and double support phases, which compensates for the reduced stride length. NMSM: Combining GA and NMSM enables one to track disease progression with enhanced precision. This has been demonstrated across a range of neuromuscular pathologies and healthy individuals. Critically, for each individual a neuromusculoskeletal model is created, driven by the individual's own EMG signals, and tracking their biomechanics, as has recently applied in neurologically impaired individuals. This creates a novel model that links in vivo neuromuscular functions to the individual, thus providing new biomarkers to assess and track PD motor impairment. Furthermore, since joint stiffness depends both on neural recruitment and mechanical properties, it is likely to provide a potent representation of neural and musculoskeletal PD impairment. Significance and potential impact This project addresses the potential for OWET to restore normal gait in PD patients. OWET aims to improve overall body motion and lower joint stiffness in those with PD, thereby improving function, quality of life, and reducing risk of injurious falls. The proposed robotic device (Ekso GT™, EksoBionics, Richmond, CA, USA) relies functions by providing passive assistance to the ankle joint, which affects the rest of the body through mechanical coupling. Currently, the amount of device assistance is estimated based on a therapist experience and expertise. Modern motion analysis methods enable us to objectively assess the required assistance providing a means to tailor the assistance to each individual and remove the risks of clinical guesswork. Robotic devices assist the physical therapist by providing task-specific repeatable mechanical action to support therapies and enable higher intensity of training. OWET aims to reduce lower limb joint stiffness, which is a recognized biomarker of PD, thereby enhancing rehabilitation for PD patients. Findings linked with the proposed study will likely give substantial solutions to the management of gait and postural disorders (posture, balance, and gait) in PD where valid interventions (pharmacological, neurosurgery, traditional physiotherapy) are lacking. Moreover, a NMSM that identifies patient-specific variables for therapy could be used to assess treatment outcomes but also to conduct on-line rehabilitation therapy by means remote control of the assistive device. This will provide a number of advantages over conventional approaches proposing an active treatment that is personalized and scalable to large populations and including a standardized training environment and an adaptable support that has the ability to increase the treatment intensity and dose, without being a burden on therapists. OWET is thus an ideal means to complete conventional therapy in clinic, while rehabilitation robots bear the potential for continued home therapy using simpler devices. Study design The study will be carried out over 36 months. Patients with clinically established diagnosis of PD according to the U.K. Parkinson's Disease Society Brain Bank Diagnostic criteria will be recruited. The diagnosis will be reviewed by a neurologist specialized on movement disorders. Briefly, 50 patients with mild to moderate disease severity, will be enrolled according to the inclusion\exclusion criteria included in the dedicated section below. The activities will be organized into 4 work packages (WP), each with measurable outputs verified by scheduled deliverables/milestones. WP1: Clinical Trial. The sample size has been defined based on published data of gait velocity in PD subjects, and a target sample of fifty PD individuals (divided in two separate cohorts) has been selected to achieve a power of at least 80% for detecting a mean group differences in mean gait velocity (p = 0.05). One cohort (n=25) will undergo a multidisciplinary intensive rehabilitation treatment, and the other will be treated with OWET. At baseline (T0) subjects will undergo neurophysiological evaluation (EEG-fMRI) and GA. Participants will then undergo an 8-weeks OWET. After the therapy (T1), the subjects will be evaluated, with the same protocol as at T0. After another 2 months, a follow-up (T2) will be conducted using the same protocol as T0. WP2: Motion analysis. State-of-the-art posture and GA will be performed pre- and post-rehabilitation. WP3: NMSM. Using the data collected in WP1 and WP2, NMSM will be performed to obtain muscular force and joint stiffness. This NMSM will be used to assess PD neuromuscular function pre- and post-rehabilitation. WP4: Neurophysiological assessment. A 256 channel High Density EEG (HD-EEG) recordings and analysis will be used to assess brain oscillation activity changes before and after the treatment. Multimodal brain imaging will be performed by simultaneous acquisition and analysis of neurophysiological signals (EEG/EMG) and fMRI data to assess the resting state connectivity and activation differences between pre- and post-treatment and to identify if the changes in the cortical activity is linked with the changes detected in WP3. Anticipated Results Locomotor functions are positively recovered by a functional gait training. Indeed, in post-stroke subjects, patients who underwent this therapy have already shown to be more likely to achieve an independent walking than people who did not receive the same treatment. OWET will improve quality of gait and balance. Effects with OWET will impact the quality of life. The results with OWET will provide innovative information for rehabilitative programs. The impact of the intervention will be assessed by measurable outcomes listed in the dedicated section below.

Device: EksoGT. EksoGT is an overground wearable gait trainer. The therapy will be carried out 3 days a week for 4 weeks.

Device: No device. The functional kinematic training will be delivered as comparator treatment and will be carried out 3 days a week for 4 weeks.

EksoGT is an overground wearable gait trainer. The therapy will be carried out 3 days a week for 4 weeks.

Device: No device. The functional kinematic training will be delivered as comparator treatment and will be carried out 3 days a week for 4 weeks.

Joint kinematics (degrees): trunk, pelvis, hip, knee, ankle (flexion-extension, ab-adduction, internal - external rotation)

Joint kinematics (degrees): trunk, pelvis, hip, knee, ankle (flexion-extension, ab-adduction, internal - external rotation)

Step duration (seconds), gait period (seconds),stance period (seconds), swing period (seconds), double support (seconds)

Step duration (seconds), gait period (seconds),stance period (seconds), swing period (seconds), double support (seconds)

Balance during Romberg Test. From the center of pressure (COP) the following parameters will be extracted: mean distance from centre of COP trajectory (mm), root mean square of COP time series (mm), sway path, total COP trajectory length (mm), range of COP displacement (mm).

Balance during Romberg Test. From the center of pressure (COP) the following parameters will be extracted: mean distance from centre of COP trajectory (mm), root mean square of COP time series (mm), sway path, total COP trajectory length (mm), range of COP displacement (mm)

Balance during Romberg Test. From the center of pressure (COP) the following parameters will be extracted: mean COP velocity (mm/s)

Balance during Romberg Test. From the center of pressure (COP) the following parameters will be extracted: mean COP velocity (mm/s)

Balance during Romberg Test. From the center of pressure (COP) the following parameters will be extracted: mean frequency (Hz), i.e., number, per second, of loops that have to be run by COP to cover total trajectory equal to sway path ; median frequency (Hz), frequency below which 50% of total power is present; 95% power frequency (Hz), frequency below which 95% of total power is present, centroidal frequency (Hz), frequency at which spectral mass is concentrated.

Balance during Romberg Test. From the center of pressure (COP) the following parameters will be extracted: mean frequency (Hz), i.e., number, per second, of loops that have to be run by COP to cover total trajectory equal to sway path ; median frequency (Hz), frequency below which 50% of total power is present; 95% power frequency (Hz), frequency below which 95% of total power is present, centroidal frequency (Hz), frequency at which spectral mass is concentrated.

Balance during Romberg Test. From the center of pressure (COP) the following parameters will be extracted: area of 95% confidence circumference (mm^2), area of 95% confidence ellipse (mm^2).

Balance during Romberg Test. From the center of pressure (COP) the following parameters will be extracted: area of 95% confidence circumference (mm^2), area of 95% confidence ellipse (mm^2).

Balance during Romberg Test. From the center of pressure (COP) the following parameters will be extracted: sway area, computed as area included in COP displacement per unit of time (mm^2/seconds).

Balance during Romberg Test. From the center of pressure (COP) the following parameters will be extracted: sway area, computed as area included in COP displacement per unit of time (mm^2/seconds).

Change in Movement Disorder Society - Unified Parkinson Disease Rating Scale (MDS-UPDRS) after 30 days

Change in Movement Disorder Society - Unified Parkinson Disease Rating Scale (MDS-UPDRS) after 60 days

Ziegler Protocol for the assessment of FOG severity (0 no festination, no FOG - 1 festination - 2 FOG).

Ziegler Protocol for the assessment of FOG severity (0 no festination, no FOG - 1 festination - 2 FOG).

The New Freezing of Gait Questionnaire (N-FOGQ) (0 never happened, 4 unable to walk for more than 30s).

The New Freezing of Gait Questionnaire (N-FOGQ) (0 never happened, 4 unable to walk for more than 30s).

Inclusion Criteria: Patient with rigid-acinetic bilateral PD form Hoehn-Yahr between 3-4 At least 4 years of disease history Stable drug therapy response without any change performed in the 3 months before the study Presence of freezing (FOG) and of postural instability not responding to parkinsonian therapy Mini Mental State Evaluation > 24/30 Exclusion Criteria: Systemic illness Presence of cardiac pacemaker Postural abnormalities, orthopedic comorbidities that do not match the active physiotherapy treatment Presence of deep brain stimulation Presence of severe disautonomia with marked hypotension Obsessive-Compulsive disorder (OCD) Major depression Dementia and psychosis History or active neoplasia Pregnancy Other criteria that do not respect the device counterindications

Benedetti MG, Beghi E, De Tanti A, Cappozzo A, Basaglia N, Cutti AG, Cereatti A, Stagni R, Verdini F, Manca M, Fantozzi S, Mazza C, Camomilla V, Campanini I, Castagna A, Cavazzuti L, Del Maestro M, Croce UD, Gasperi M, Leo T, Marchi P, Petrarca M, Piccinini L, Rabuffetti M, Ravaschio A, Sawacha Z, Spolaor F, Tesio L, Vannozzi G, Visintin I, Ferrarin M. SIAMOC position paper on gait analysis in clinical practice: General requirements, methods and appropriateness. Results of an Italian consensus conference. Gait Posture. 2017 Oct;58:252-260. doi: 10.1016/j.gaitpost.2017.08.003. Epub 2017 Aug 5.

Sartori M, Fernandez JW, Modenese L, Carty CP, Barber LA, Oberhofer K, Zhang J, Handsfield GG, Stott NS, Besier TF, Farina D, Lloyd DG. Toward modeling locomotion using electromyography-informed 3D models: application to cerebral palsy. Wiley Interdiscip Rev Syst Biol Med. 2017 Mar;9(2). doi: 10.1002/wsbm.1368. Epub 2016 Dec 21.

Bortole M, Venkatakrishnan A, Zhu F, Moreno JC, Francisco GE, Pons JL, Contreras-Vidal JL. The H2 robotic exoskeleton for gait rehabilitation after stroke: early findings from a clinical study. J Neuroeng Rehabil. 2015 Jun 17;12:54. doi: 10.1186/s12984-015-0048-y.

Volpe D, Signorini M, Marchetto A, Lynch T, Morris ME. A comparison of Irish set dancing and exercises for people with Parkinson's disease: a phase II feasibility study. BMC Geriatr. 2013 Jun 4;13:54. doi: 10.1186/1471-2318-13-54.

Sartori M, Lloyd DG, Farina D. Corrections to "Neural Data-Driven Musculoskeletal Modeling for Personalized Neurorehabilitation Technologies". IEEE Trans Biomed Eng. 2016 Jun;63(6):1341. doi: 10.1109/TBME.2016.2563138.

Galli M, Cimolin V, De Pandis MF, Le Pera D, Sova I, Albertini G, Stocchi F, Franceschini M. Robot-assisted gait training versus treadmill training in patients with Parkinson's disease: a kinematic evaluation with gait profile score. Funct Neurol. 2016 Jul-Sep;31(3):163-70. doi: 10.11138/fneur/2016.31.3.163.

Sale P, De Pandis MF, Le Pera D, Sova I, Cimolin V, Ancillao A, Albertini G, Galli M, Stocchi F, Franceschini M. Robot-assisted walking training for individuals with Parkinson's disease: a pilot randomized controlled trial. BMC Neurol. 2013 May 24;13:50. doi: 10.1186/1471-2377-13-50.

Fundaro C, Giardini A, Maestri R, Traversoni S, Bartolo M, Casale R. Motor and psychosocial impact of robot-assisted gait training in a real-world rehabilitation setting: A pilot study. PLoS One. 2018 Feb 14;13(2):e0191894. doi: 10.1371/journal.pone.0191894. eCollection 2018.

Picelli A, Melotti C, Origano F, Waldner A, Fiaschi A, Santilli V, Smania N. Robot-assisted gait training in patients with Parkinson disease: a randomized controlled trial. Neurorehabil Neural Repair. 2012 May;26(4):353-61. doi: 10.1177/1545968311424417. Epub 2012 Jan 18.

Gassert R, Dietz V. Rehabilitation robots for the treatment of sensorimotor deficits: a neurophysiological perspective. J Neuroeng Rehabil. 2018 Jun 5;15(1):46. doi: 10.1186/s12984-018-0383-x.

Del Din S, Bertoldo A, Sawacha Z, Jonsdottir J, Rabuffetti M, Cobelli C, Ferrarin M. Assessment of biofeedback rehabilitation in post-stroke patients combining fMRI and gait analysis: a case study. J Neuroeng Rehabil. 2014 Apr 9;11:53. doi: 10.1186/1743-0003-11-53.

Scarton A, Jonkers I, Guiotto A, Spolaor F, Guarneri G, Avogaro A, Cobelli C, Sawacha Z. Comparison of lower limb muscle strength between diabetic neuropathic and healthy subjects using OpenSim. Gait Posture. 2017 Oct;58:194-200. doi: 10.1016/j.gaitpost.2017.07.117. Epub 2017 Jul 31.

McGrath RL, Pires-Fernandes M, Knarr B, Higginson JS, Sergi F. Toward goal-oriented robotic gait training: The effect of gait speed and stride length on lower extremity joint torques. IEEE Int Conf Rehabil Robot. 2017 Jul;2017:270-275. doi: 10.1109/ICORR.2017.8009258.

Volpe D, Pavan D, Morris M, Guiotto A, Iansek R, Fortuna S, Frazzitta G, Sawacha Z. Underwater gait analysis in Parkinson's disease. Gait Posture. 2017 Feb;52:87-94. doi: 10.1016/j.gaitpost.2016.11.019. Epub 2016 Nov 10.

Frazzitta G, Balbi P, Maestri R, Bertotti G, Boveri N, Pezzoli G. The beneficial role of intensive exercise on Parkinson disease progression. Am J Phys Med Rehabil. 2013 Jun;92(6):523-32. doi: 10.1097/PHM.0b013e31828cd254.

Durandau G, Farina D, Sartori M. Robust Real-Time Musculoskeletal Modeling Driven by Electromyograms. IEEE Trans Biomed Eng. 2018 Mar;65(3):556-564. doi: 10.1109/TBME.2017.2704085. Epub 2017 May 12.

Sartori M, Maculan M, Pizzolato C, Reggiani M, Farina D. Modeling and simulating the neuromuscular mechanisms regulating ankle and knee joint stiffness during human locomotion. J Neurophysiol. 2015 Oct;114(4):2509-27. doi: 10.1152/jn.00989.2014. Epub 2015 Aug 5.

Sawacha Z, Carraro E, Del Din S, Guiotto A, Bonaldo L, Punzi L, Cobelli C, Masiero S. Biomechanical assessment of balance and posture in subjects with ankylosing spondylitis. J Neuroeng Rehabil. 2012 Aug 29;9:63. doi: 10.1186/1743-0003-9-63.

Sawacha Z, Guarneri G, Cristoferi G, Guiotto A, Avogaro A, Cobelli C. Integrated kinematics-kinetics-plantar pressure data analysis: a useful tool for characterizing diabetic foot biomechanics. Gait Posture. 2012 May;36(1):20-6. doi: 10.1016/j.gaitpost.2011.12.007. Epub 2012 Mar 30.

Teramoto H, Morita A, Ninomiya S, Akimoto T, Shiota H, Kamei S. Relation between Resting State Front-Parietal EEG Coherence and Executive Function in Parkinson's Disease. Biomed Res Int. 2016;2016:2845754. doi: 10.1155/2016/2845754. Epub 2016 Jun 28.

Herz DM, Eickhoff SB, Lokkegaard A, Siebner HR. Functional neuroimaging of motor control in Parkinson's disease: a meta-analysis. Hum Brain Mapp. 2014 Jul;35(7):3227-37. doi: 10.1002/hbm.22397. Epub 2013 Oct 5.

Pizzolato C, Lloyd DG, Sartori M, Ceseracciu E, Besier TF, Fregly BJ, Reggiani M. CEINMS: A toolbox to investigate the influence of different neural control solutions on the prediction of muscle excitation and joint moments during dynamic motor tasks. J Biomech. 2015 Nov 5;48(14):3929-36. doi: 10.1016/j.jbiomech.2015.09.021. Epub 2015 Oct 19.

Dennis R. Louie, Powered robotic exoskeletons in post-stroke rehabilitation of gait: a scoping review

A De Luca, Exoskeleton for Gait Rehabilitation: Effects of Assistance, Mechanical Structure, and Walking Aids on Muscle Activations