Brain Computer Interface (BCI) Technology for Stroke Hand Rehabilitation (ARTS-BCI)

ARTS-BCI: Advanced Brain Computer Interface (BCI) Technology for Wrist and Hand Rehabilitation After Stroke

This study is carried out to find out if Brain Computer Interface (BCI) technology or BCI technology coupled with robotic technology using a Haptic Knob will benefit patients with arm paralysis after stroke. BCI uses EEG-based motor imagery to detect user's thinking abilities which control motor movement. Haptic Knob is a novel robotic device, which specifically trains the wrist and hand with intensive repetitions in a supported environment.

Physical therapy approaches are the de facto rehabilitation for stroke, which involve human therapists to assist stroke patients in recovering their motor ability. Modern rehabilitation technologies include robotics, functional electrical stimulation, transcranial magnetic stimulation and virtual reality. Robotic rehabilitation alleviates the labor-intensive aspects of physical rehabilitation by human therapists and could potentially improve the productivity of stroke rehabilitation. However, it is fundamentally based on movement repetition with visual feedback that helps stroke patients improve motor ability in their weak stroke-affected arms and legs. However, the robot is still able to move the weak part of the patient even if the patient is not attentive towards the training and thus the robotic training becomes a passive activity. In contrast, BCI-based robotic training works by ensuring active engagement by the hemiparetic patients in making a volitional movement. In addition, hemiplegic or locked-in stroke patients who do not have any motor power on the affected limbs are then able to engage and perform a volitional movement on these affected limbs. BCI-based robotic rehabilitation fills this gap by detecting the motor intent of hemiplegic patients from the Electroencephalogram (EEG) signals to drive the robotic rehabilitation. This BCI-based robotic rehabilitation for stroke research project was jointly conducted by Tan Tock Seng Hospital (TTSH), National Neuroscience Institute (NNI) and Institute for Infocomm Research (I2R). Preliminary clinical trials performed at TTSH have shown that stroke patients can operate the BCI as effective as healthy subjects. Specifically, this research project will address the following gaps in the area of rehabilitation for stroke: Single-modal BCI - The current system employs a single modal non-invasive EEG-based BCI that detects motor intent using at least 2.5 seconds of EEG data. Hence, the research of an advanced multi-modal BCI such as synergizing near-infrared spectroscopy with EEG to yield a more responsive and effective BCI-based robotic rehabilitation system is proposed. Standard therapy - The current system employs a standard therapy for all the stroke patients. However, physiotherapists and occupational therapists usually adopt a more individualized therapy for each stroke patients. Hence, research on an individualized therapy for each stroke patient according to his or her learning rate and neurological insult is proposed. Only physiological rehabilitation - The current system only performs physiological rehabilitation of motor functions of stroke patients. Currently some validated scales for post-stroke depression such as Beck depression inventory, CES-D, Zung scale, State trait, HADS etc are difficult to administer in stroke patients who cannot participate with assessment due to impaired language or cognitive abilities. Hence an advanced BCI-based rehabilitation system that also detects the mental state of the stroke patient is proposed to cover both physiological and psychological rehabilitation. Upper Limb rehabilitation - The current system which uses the clinically-proven MIT Manus robotic rehabilitation system, only performs upper limb rehabilitation for stroke patients in gross reach patterns. Human hand skills, in contrast, consist of more complex manipulation movement patterns which can be intervened by BCI-based robotic rehabilitation. Hence, an advanced BCI-based rehabilitation system that covers the hand function is proposed to cover the rehabilitation of the entire upper extremity.

Stroke Rehabilitation, Brain Computer Interface, Upper Limb Rehabilitation, EEG-based Motor Imagery, Robotic Rehabilitation, Haptic Knob

Use of conventional manual facilitation and function-based training used in conventional occupational therapy training for post-stroke upper limb weakness. Training is modelled along the neurodevelopmental techniques and will include stretching, tone management, weight bearing exercises, movement facilitation, selfcare training, arm ergometry by arm bicycles and grip strength training. Training intensity is 1.5 hours for 3 times a week for 6 weeks consecutively.

BCI based robotic rehabilitation works by detecting the motor intent of the user from electroencephalogram signals to drive the robotic rehabilitation via Haptic Knob. Training intensity is 1.5 hours for 3 times a week for 6 weeks consecutively.

Haptic Knob is an upper limb robot designed for use in robotic-assisted rehabilitation of the stroke wrist and hand. Training intensity is 1.5 hours for 3 times a week for 6 weeks consecutively.

Measurements at 2 weeks prior to intervention, just before start of intervention, at midpoint of intervention, at completion of intervention, at 3 months post intervention and at 6 months post-intervetion

Measurements at 2 weeks prior to intervention, just before start of intervention, at midpoint of intervention, at completion of intervention, at 3 months post intervention and at 6 months post-intervetion

Measurements at 2 weeks prior to intervention, just before start of intervention, at midpoint of intervention, at completion of intervention, at 3 months post intervention and at 6 months post-intervetion

Measurements at 2 weeks prior to intervention, just before start of intervention, at midpoint of intervention, at completion of intervention, at 3 months post intervention and at 6 months post-intervetion

Measurements at 2 weeks prior to intervention, just before start of intervention, at midpoint of intervention, at completion of intervention, at 3 months post intervention and at 6 months post-intervetion

Measurements at 2 weeks prior to intervention, just before start of intervention, at midpoint of intervention, at completion of intervention, at 3 months post intervention and at 6 months post-intervetion

Measurements at 2 weeks prior to intervention, just before start of intervention, at midpoint of intervention, at completion of intervention, at 3 months post intervention and at 6 months post-intervetion

Inclusion Criteria: Aged 21-80 years with first-ever clinical stroke, within 1-24 months onset. Stroke type: ischemic or haemorhagic. Fugl-Meyer motor score of the upper limb range from 10-50 or Motor power MRC grade 3-5 in shoulder abductors and elbow flexors, and 0-3 in wrist dorsiflexors and finger flexors Ability to pay attention and maintain supported sitting for 1 hour continuously. Able to give own consent and understand simple instructions Fulfills BCI and Haptic knob physical screening trial. Exclusion Criteria: Functional status: severe aphasia or inattention, unstable medical conditions which may affect participation (e.g. unresolved sepsis, postural hypotension, end stage renal failure) or anticipated life expectancy of <1 year due to malignancy or neurodegenerative disorder) Hemispatial neglect (visual or sensory) or severe visual impairment despite visual aids. Epilepsy, severe depression or psychiatric disorder. Recurrent stroke Skull defect as this would affect physical fit of EEG cap interface. Local arm factors: Severe spasticity Modified Ashworth scale >2 in any region, visual analogue scale (VAS score) >4/10, fixed joint contracture , patients with poor skin conditions, infections or eczema which may potentially be worsened by robotic shell contact.

Ang KK, Guan C, Chua KS, Ang BT, Kuah C, Wang C, Phua KS, Chin ZY, Zhang H. Clinical study of neurorehabilitation in stroke using EEG-based motor imagery brain-computer interface with robotic feedback. Annu Int Conf IEEE Eng Med Biol Soc. 2010;2010:5549-52. doi: 10.1109/IEMBS.2010.5626782.

Arvaneh M, Guan C, Ang KK, Ward TE, Chua KSG, Kuah CWK, Ephraim Joseph GJ, Phua KS, Wang C. Facilitating motor imagery-based brain-computer interface for stroke patients using passive movement. Neural Comput Appl. 2017;28(11):3259-3272. doi: 10.1007/s00521-016-2234-7. Epub 2016 Mar 4.

Ang KK, Guan C, Phua KS, Wang C, Zhou L, Tang KY, Ephraim Joseph GJ, Kuah CW, Chua KS. Brain-computer interface-based robotic end effector system for wrist and hand rehabilitation: results of a three-armed randomized controlled trial for chronic stroke. Front Neuroeng. 2014 Jul 29;7:30. doi: 10.3389/fneng.2014.00030. eCollection 2014.

Ang KK, Guan C, Chua KS, Phua KS, Wang C, Chin ZY, Zhou L, Tang KY, Joseph GJ, Kuah C. A clinical study of motor imagery BCI performance in stroke by including calibration data from passive movement. Annu Int Conf IEEE Eng Med Biol Soc. 2013;2013:6603-6. doi: 10.1109/EMBC.2013.6611069.