A new lower-limb training system is introduced to enhance the clinical service for post-stroke lower limb rehabilitation and to assist the establishment of public clinical trial in different settings and share experiences on the robot-assisted functional training.
Stroke is caused by intracranial haemorrhage or thrombosis, which cuts off arterial supply to brain tissue and usually damages the motor pathway of the central nervous system affecting one side of the body. Reduced descending neural drive to the affected side could lead to hemiplegia, which significantly influences the activity of daily living (ADL) of stroke survivors (Singam, Ytterberg, Tham \& von Koch, 2015). While the upper-limb motor impairment could be compensated using the contralateral side for picking up or manipulating objects, the loss of motor functionality on the lower limb would substantially limit the mobility and body balance. Many stroke survivors are dependent on walking aids or manual support from caregivers for standing and walking, otherwise they would have great risk of falling with serious consequences (Tasseel-Ponche, Yelnik \& Bonan, 2015). Recent studies suggest stroke patients could relearn walking ability by developing alternative neural circuitries through long-term adaptation process, known as neuroplasticity. High-intensity, repetitive, and task-specific gait training is the key to enhance gait recovery of hemiplegic stroke patients (Kreisei, Hennerici \& Bäzner, 2007; Kleim \& Jones, 2008). The development of robot-assisted lower-limb exoskeleton devices has great clinical potential in stroke rehabilitation. Many lower-limb exoskeleton robots are clinically-available for non-ambulatory stroke patients to practice walking with passive assistance on body-weight-supported treadmill training (BWSTT) (Morone, et al., 2017). Existing robot-assisted gait training (RAGT) such as Lokomat and electromechanical Gait Trainer provide automatic, rhythmic, and repetitive powered assistance to major lower-limb joints at hips and knees bilaterally (Poli, Morone, Rosati \& Masiero, 2013). Large-scale randomized controlled trials (RCT) of these RAGT in combination with conventional therapies show significantly more chronic stroke patients improved functional gait independency and ADL than receiving conventional therapies alone (Pohl, et al., 2007; Schwartz, et al., 2009; Hidler, et al., 2009; Mehrholz, et al., 2013). However, Hesse, Schmidt, Werner \& Bardeleben (2003) suggest the integration of robots into gait rehabilitation could merely be an auxiliary tool for therapists to enhance training intensity and safety without increasing their workload. Most clinically-available RAGT are bounded to treadmill with passive assistance (van Peppen, et al., 2004; Morone, et al., 2017), but researches show task-variations and active participation in gait training could improve retention of newly-learnt skills and could promote generalization of training effects (Salbach, et al., 2004; Kwon, Woo, Lee \& Kim, 2015). Portable RAGT that allows active over-ground gait training would be more promising especially for ambulatory stroke patients. Robot-assisted ankle foot orthosis (AFO) and knee brace are good candidates of portable exoskeleton devices for RAGT of hemiplegic stroke patients (Duerinck, et al., 2012; Zhang, Davies \& Xie, 2013; Mehrholz, et al., 2017). Conventional AFO is mainly designed for treating foot drop gait abnormality with passive support in ankle dorsiflexion for foot clearance in swing phase and shock absorption in loading response. Conventional knee brace is mainly designed for body support in stance phase. The integration of robot assistance in the affected ankle and/or knee joint could provide active power assistance that synchronises to patients' voluntary residual ankle and/or knee movement. Long-term active power assistance might stimulate experience-driven gait recovery or develop compensatory gait pattern to facilitate gait (Kleim \& Jones, 2008). In order to translate robotic rehabilitation research into clinical application, evidence-based clinical research should be carried out to test the safety and effectiveness of the new devices or interventions on stroke patients (Backus, Winchester \& Tefertiller, 2010). Many designs of robot-assisted AFO and knee braces have been proposed by different research groups, but most of them reported only the results of feasibility tests, mainly on healthy subjects with small sample sizes (Dollar \& Herr, 2008; Shorter, et al., 2013; Alam, Choudhury \& Bin Mamat, 2014). Majority of previous studies concerned about the immediate effects of wearing the robot-assisted AFOs and knee braces during walking, but few studies investigated the long-term therapeutic effects of wearing the devices for RAGT of stroke patients (Lo, 2012). In particular, systematic review by Mehrholz, et al. (2017) shows only one RCT has evaluated the efficacy of ankle training using robot-assisted AFO but in seated position, no RCT evaluated gait training using robot-assisted AFO on both over-ground walking and stair ambulation. In this study, the Exoskeleton Ankle Robot and Knee Robot have been proposed and evaluated as a robot-assisted AFO and knee brace for gait training of stroke patients with foot drop gait abnormality. Clinical application of robot-assisted AFO and knee brace on stroke patients has to overcome some important challenges, such as to reduce weight loading on the leg, and to achieve portability and adaptability to various walking environments. The Exoskeleton Ankle Robot and Knee Brace aims: (1) to provide synchronised active ankle and/or knee power assistance to facilitate walking, (2) to develop accurate and reliable method to classify user walking intention in over-ground walking and stair ambulation, (3) to deliver training protocol for RAGT of stroke patients with foot drop gait abnormality. The feasibility tests and RCT of the Exoskeleton Ankle Robot and Knee Brace could validate the clinical value of this new rehabilitation robot, and could potentially establish a new intervention of gait rehabilitation for stroke patients.
Study Type
INTERVENTIONAL
Allocation
RANDOMIZED
Purpose
TREATMENT
Masking
DOUBLE
Enrollment
64
Patients will wear the robotic ankle system and undergo 20-minute over-ground walking and 10-minute stair walking.
Patients will wear the robotic knee system and undergo 20-minute over-ground walking and 10-minute stair walking.
Department of Biomedical Engineering, The Chinese University of Hong Kong
Hong Kong, Hong Kong
RECRUITINGFunctional Ambulatory Category (FAC)
Functional Ambulatory Category (FAC) is a reliable measurement of independent walking ability on level-ground walking and stair ambulation, which is a good prediction of independent community walking post-stroke (Mehrholz, et al., 2007). FAC consists of 6-level scale: patients with FAC=4 requires supervision in level ground walking, FAC=5 requires supervision only when walking on non-level surface such as stairs.
Time frame: Baseline, Post-Training, 3-month follow up
Fugl-Meyer Assessment for Lower-Extremity (FMA-LE)
Fugl-Meyer Assessment for Lower-Extremity (FMA-LE), consists of 34-level cumulative scoring system to examine lower-limb functions of hemiplegic stroke patients quantitatively through a set of lower-limb movement tasks in reflex, flexor/extensor synergy, volitional movement, coordination and speed (Fugl-Meyer, et al., 1975). All assessment items are either scoring "full", "partial", or "none" functionality in the affected side, which minimizes ceiling and floor effects. FMA-LE demonstrated high internal consistency and a reliable assessment tool for a group of 140 hemiplegic community dwelling patients (Park \& Choi, 2014).
Time frame: Baseline, Post-Training, 3-month follow up
Modified Ashworth Scale (MAS)
Modified Ashworth Scale (MAS), consists of 4-level scale to examine joint spasticity based on muscle tone and resistance detected during passive stretching with good inter-rater reliability (ICC =0.85) (Bohannon \& Smith, 1987).
Time frame: Baseline, Post-Training, 3-month follow up
Berg Balance Scale (BBS)
Berg Balance Scale (BBS), consists of 56-level measures to examine balance ability and to predict falling risk with high reliability (ICC=0.98) (Steffen, Hacker \& Mollinger, 2002). Stroke patients were assessed based on their performance on 14 simple mobility tasks, including transfer, standing, and reaching.
Time frame: Baseline, Post-Training, 3-month follow up
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.
Timed 10-Meter Walk Test (10mWT)
Timed 10-Meter Walk Test (10mWT), measures comfortable and fast walking speeds in short distance. The ability to increase walking speed above a comfortable pace suggests the capability to adapt to varying environments, such as crossing street, with high reliability (ICC=0.90-0.96) (Flansbjer, et al., 2005). Average walking speed of healthy elderly subjects ranges in 0.6m/s-1.4m/s, and can increase to 21%-56% above the comfortable pace for faster walking speed.
Time frame: Baseline, Post-Training, 3-month follow up
6-minute Walk Test (SMWT)
Six-Minute Walk Test (SMWT), measures the maximum walking distance covered in fixed duration as a sub-maximal test of endurance and aerobic capacity. The measurement of 6MWT is highly correlated to FAC (Mehrholz, et al., 2007) with good reliability (ICC=0.94-0.96) (Steffen, Hacker \& Mollinger, 2002).
Time frame: Baseline, Post-Training, 3-month follow up