For patients whose stroke affected their ability to use their hand, a new electrical stimulation device may help. The device allows patients to control their impaired hand using their unaffected hand, and to control the timing and intensity of electrical stimulation. In a study published online on September 8 in Stroke, the new method led to improvements in hand dexterity.
Patient-Controlled Electrical Stimulation
Electrical stimulation is already widely available and used in stroke rehabilitation to help patients recover the use of their limbs. Traditional electrical stimulation, known as cyclic neuromuscular electric stimulation (cNMES), is controlled by a therapist and requires no active participation from the patient. By contrast, the new method, called contralaterally controlled functional electrical stimulation or CCFES, allows patients to exercise an impaired hand by controlling electrical stimulation to the impaired hand using their strong hand (or contralateral hand) and performing tasks.
The researchers hypothesized that because CCFES allows patients to use both hands, is done in real time, and involves tasks, it may result in better, faster rehabilitation compared to cNMES.
Researchers at MetroHealth Medical Center in Cleveland, Ohio enrolled 80 patients who had had a stroke and were partially paralyzed in one of their upper limbs for at least six weeks. Half the patients received 10 sessions per week of the new stimulation method, and half received 10 sessions of the traditional stimulation, for six months.
After the six months, the researchers administered the Box and Block Test (BBT) to gauge improvements in hand dexterity. The test counts how many times a person can pick up a block, move it over a partition, and release it in a target area within 60 seconds.
Patients who received CCFES had greater improvement in dexterity as measured by the Box and Block Test than patients who underwent cNMES. Patients who’d had a stroke less than two years prior to the study and had moderate, not severe, hand impairment at the start of the study had the biggest improvements.
A New Option for Stroke Rehabilitation
For recent stroke patients with moderate hand impairment, CCFES is a better option than cNMES, the study authors say, possibly because CCFES happens in real time, requires patients to open both hands at once, and/or allows patients to practice tasks with the impaired hand. They added that with CCFES—unlike with cNMES—patients can control the timing and intensity of tasks.
Watch a video showing the new stimulation device in action here:
Uzma Samadani, MD, PhD, FAANS; Samuel R. Daly | Features
AANS Neurosurgeon: Volume 25, Number 3, 2016
TBI is the leading cause of death and disability in Americans under age 35 and the leading cause of premature death and disability worldwide (1). In 2009 alone, 2.4 million patients presented to an emergency department (ED) with a TBI and an additional 52,695 died from their injury (2). Currently, at least 5.3 million U.S. residents live with long-term disabilities related to a TBI (3). The economic toll of TBI has been reported as $75 billion for a single year (4). Between 2002 and 2010, the rate of TBI in the U.S. population increased by over 50 percent, (1) excluding the military and those that do not seek medical care (5).
Despite the colossal nature of the problem, little progress has been made in developing new therapeutics for TBI. A PubMed search reveals 30 failed clinical trials for TBI since 1993, 25 of which have been in the last 15 years and 13 of which have been in the last five years (6-35). These 30 trials failed to find a significant effect for treatments that were supported by extensive preclinical studies and Phase I and II trials. They included hypothermia and temperature control (11 studies), hypertonic saline (three studies), progesterone (two studies), prostacyclin (two studies), surgical intervention (one study), intracranial pressure monitoring (one study) and a number of other pharmacological interventions (10 studies). The estimated total cost of these trials is $1.1 billion (36).
All 30 of these trials were prospective, randomized, controlled studies based on well-executed preliminary studies. Ninety-three percent indicated blinded assessment, and 70 percent were done at multiple centers. Patient retention was unprecedented. For example, the SYNAPSE trial tested the effect of progesterone in a double-blind, randomized manner with more than 1,000 participants and 96 percent, six-month retention. Thus, rather than blaming standard methodological limitations for these failed trials, one must look deeper into how TBI trials recruit and assess the outcome of the patients.
Table 1: Major Inclusion Criteria
Number of Studies
GCS + CT
GCS + Pupil Reactivity
GCS + ICP
The Glasgow Coma Scale (GCS) was used as a major inclusion criterion in 29 of the 30 studies (97 percent) and was the only major inclusion criteria in 18 studies (Table 1). This measure, which has been in use for 40 years, conveys the clinically-relevant acute exam of the verbal, motor and eye movement response of a patient. Using GCS to select patients for clinical trials may be suboptimal because it does not account for diverse pathophysiology at any level of brain injury severity. For example, a patient with severe TBI (GCS 3-8) can be minimally or non-responsive due to a wide diversity of underlying pathophysiology with different outcomes such as a non-significant impact while intoxicated, a treatable subdural or epidural hematoma or a diffuse axonal or anoxic injury, the latter of which may be associated with very poor outcomes. Recognition of the limitations of the mild/moderate/severe paradigm was the impetus for the 3,000-patient, 13-center TRACK-TBI study (37) and the European CENTER TBI project (38).
Imaging was only used as a major inclusion or exclusion criteria in eight of the 30 failed trials; however, conventional acute imaging performed for trauma may not fully differentiate complex pathophysiology, such as diffuse axonal injury (DAI) or anoxic injury.
Table 2: Outcome Measures Utilized
Number of Studies
GOS(-E) as Primary Outcome
GOS(-E) as Secondary Outcome
Length of Stay
The 5-point Glasgow Outcome Scale (GOS), or its extended 8-point version (GOS-E) was used as the primary outcome measure in 21 of the 30 studies (70 percent), and was the only outcome measure used in six of those studies (Table 2). Four of the remaining studies used the GOS as a secondary outcome measure, and two used an outcome assessment that is structured similarly to the GOS in the pediatric populations (Pediatric Cerebral Performance Category). Three studies used other means of measuring outcome. While GOS or GOS-E accurately captures global phenomena, it may fail to assess subtle differences in outcomes over a wide range of functioning. We speculate that clinical trials for brain injury will have a better probability of success when there are means of detecting and classifying brain injury appropriately, according to its pathophysiology as patients enter a trial and more sensitive outcome measures to assess recovery as patients leave a trial are utilized.
Multimodal algorithmic assessment to obtain an accurate pathophysiologic diagnosis is routine with virtually every other body system. For example, when a patient presents to the emergency room with chest pain, the workup includes imaging of the heart and lungs (echocardiogram, angiogram, chest x-ray and/or CT scan), an assessment of the electrical activity of the heart (electrocardiogram) and blood tests to analyze the concentration of various proteins that indicate cardiac or other pathology, such as troponin or D-dimers. No one would ever conceive of conducting a clinical trial for “chest pain” based on only a physical examination and a single imaging study to assess the impact of a single universal intervention with an 8-point outcome measure. On the contrary, initiatives for precision medicine seek to define even baseline patient characteristics prior to accurate diagnosis. The complexity of the central nervous system even before an injury befuddles simple functional assessment. TBI should warrant a similar multimodal assessment that enables a more clear understanding of the underlying pathophysiology of the type of brain injury before enrollment in a trial.
Among the many possible assessment tools being investigated for better classification of brain injury are serum biomarkers, eye tracking and magnetic resonance imaging (MRI). Our laboratory is among many engaged in research to enable better classification of brain injury. We will prospectively enroll more than 1,000 trauma patients at Hennepin County Medical Center (HCMC) in Minneapolis and will follow these individuals for a year after their injury. This study aims to develop a multi-modal classification scheme for brain injury that will be able to accurately diagnose acute pathophysiology in brain-injured subjects with eye tracking, the analysis of protein markers in patient blood samples and radiographic imaging.
Abnormal eye movements are found in up to 90 percent of patients with so-called “mild” TBI or concussion (39-44). The eye tracking employed in this study will be non-spatially calibrated and used to detect subtle abnormalities in motility and sustained vergence; the ability of the eyes to focus on a single point in space over time (45). Eye tracking fundamentally detects palsies in cranial nerves III and VI (46), and eye tracking metrics correlate with the severity of concussion symptoms and the improvement of those symptoms over time in both an adult emergency department population (47) and in a pediatric concussion center population (48). Disruption of eye tracking metrics clinically correlates with convergence dysfunction and abnormal near point of convergence in children (48). Eye tracking detects concussion as defined by its symptoms with a high sensitivity and specificity (49).
Protein biomarkers in the serum of TBI patients have been extensively researched in the last 20 years and represent potentially promising indicators of the nature of the brain injury, including the exact type of cellular injury that has occurred in the brain (neuronal, glial or axonal).
Preliminary models have already been proposed providing evidence that can aid in determining acute treatment plans, but these models are limited by statistical power and their nature as uni-variate models of blood-based biomarkers (50-53). Combining eye-tracking data with data from biomarkers in blood samples and other clinical data (i.e. brain imaging and the physical exam) and correlating it with detailed outcome measures at a high level of statistical power into a multivariate model has great potential for improving classification of brain injuries.
A model for classification and treatment of brain injury will be beneficial not only for patient prognostication and limiting the economic toll of TBI but also in gaining a more clear understanding of the underlying mechanisms of various brain injuries. With such understanding, we will increase the probability of successful clinical trials for the treatment of TBI.
Disclosures Uzma Samadani, MD, PhD, has submitted several patents describing the eye tracking technology mentioned in this article. These patents are owned by NYU, HCMC and the VA and licensed to Oculogica Inc., a company co-founded by Dr. Samadani and in which she has an equity interest. The U.S. also has a grant from Abbott Diagnostic Laboratories to investigate serum biomarkers for brain injury.
Sertkaya, A.B., Birkenbach, A., Berlind, A., & Eyraund, J. (2014). Examination of Clinical Trial Costs and Barriers for Drug Development. Report, US Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation, Washington, DC
Samadani, U. (2015). A new tool for monitoring brain function: eye tracking goes beyond assessing attention to measuring central nervous system physiology. Neural regeneration research, 10(8), 1231-1233.
When surgery and medication don’t help people with epilepsy, electrical stimulation of the brain has been a treatment of last resort. Unfortunately, typical approaches, such as vagal nerve stimulation or responsive nerve stimulation, rarely stop seizures altogether. But a new Mayo Clinic study in JAMA Neurology shows that seizures were suppressed in patients treated with continuous electrical stimulation.
Epilepsy is a central nervous system disorder in which nerve cell activity in the brain becomes disrupted. In the study, 13 patients with drug-resistant epilepsy were deemed unsuitable for resective surgery, which removes a portion of the brain — usually about the size of a golf ball — that was causing seizures. When patients are evaluated for surgery, a grid of electrical contacts is placed on the brain to record seizures and interictal epileptiform discharges (IEDs). IEDs are electrical discharges that occur intermittently during normal brain function, and have been used as markers to locate portions of brain affected by epilepsy.
In the study, the grid of electrical contacts was used for stimulation at levels the patient would not notice. If the stimulation provided clinical benefit to the patient, this temporary grid was replaced with more permanent contacts that could offer continuous stimulation.
Ten of the 13 patients, 77 percent, reported improvement for both epilepsy severity and life satisfaction. The majority of patients experienced more than 50 percent reduction in seizures, and 44 percent were free of disabling seizures. The reduction in IED rate occurred within minutes of initiating stimulation.
«This study suggests that subthreshold cortical stimulation is both effective clinically and reduces interictal epileptiform discharges,» says lead author Brian Lundstrom, M.D., Ph.D., a neurology epilepsy fellow at Mayo Clinic. «We think this approach not only provides an effective treatment for those with focal epilepsy but will allow us to develop ways of assessing seizure likelihood for all epilepsy patients. It would be of enormous clinical benefit if we could personalize treatment regimens for individual patients without waiting for seizures to happen.»
During seizures, abnormal electrical activity in the brain sometimes results in loss of consciousness. For people with epilepsy, seizures severely limit their ability to perform tasks where even a momentary loss of consciousness could prove disastrous — driving a car, swimming or holding an infant, for example. Approximately 50 million people worldwide have epilepsy, according to the World Health Organization.
Seizures sometimes have been compared to electrical storms in the brain. Seizure signs and symptoms may include:
•A staring spell
•Uncontrollable jerking movements of the arms and legs
•Loss of consciousness or awareness
Treatment with medications or surgery can control seizures for about two-thirds of people with epilepsy. However, when drug-resistant focal epilepsy occurs in an area of the brain that controls speech, language, vision, sensation or movement, resective surgery is not an option.
«For people who have epilepsy that can’t be treated with surgery or medication, effective neurostimulation could be a wonderful treatment option,» Dr. Lundstrom says.
The risks of subthreshold cortical stimulation are relatively minimal and include typical infection and bleeding risks as well as the possibility that the stimulation would not be subthreshold and would be noticed by the patient, Dr. Lundstrom says. The authors note that further investigation is needed to quantify treatment effect and examine the effect mechanism. The authors plan to examine the efficacy of this approach in a prospective clinical trial.
This study represents ongoing efforts to restore normal function to epileptic brain tissue by using neurostimulation. Other efforts are aimed at understanding the physiologic changes that chronic stimulation produces in brain tissue.
Saebo, Inc., is a leading global provider of affordable evidenced-based therapy solutions for individuals suffering from impaired mobility and function. Headquartered in Charlotte, NC, the company was founded in 2001 by two occupational therapists specializing in upper limb recovery.
Saebo’s products and programs are now offered as a treatment option at over 2,000 clinics and hospitals nationwide, including 22 of the Top 25 Rehabilitation Hospitals as ranked by U.S. News & World Report. With a network of over 8,000 trained clinicians spanning four continents, Saebo is committed to helping clients around the globe achieve a new level of independence.
Background and Purpose— A sensory-based intervention called peripheral nerve stimulation can enhance outcomes of motor training for stroke survivors with mild-to-moderate hemiparesis. Further research is needed to establish whether this paired intervention can have benefit in cases of severe impairment (almost no active movement).
Methods— Subjects with chronic, severe poststroke hemiparesis (n=36) were randomized to receive 10 daily sessions of either active or sham stimulation (2 hours) immediately preceding intensive task-oriented training (4 hours). Upper extremity movement function was assessed using Fugl–Meyer Assessment (primary outcome measure), Wolf Motor Function Test, and Action Research Arm Test at baseline, immediately post intervention and at 1-month follow-up.
Results— Statistically significant difference between groups favored the active stimulation group on Fugl–Meyer at postintervention (95% confidence interval [CI], 1.1–6.9; P =0.008) and 1-month follow-up (95% CI, 0.6–8.3; P =0.025), Wolf Motor Function Test at postintervention (95% CI, −0.21 to −0.02; P =0.020), and Action Research Arm Test at postintervention (95% CI, 0.8–7.3; P =0.015) and 1-month follow-up (95% CI, 0.6–8.4; P =0.025). Only the active stimulation condition was associated with (1) statistically significant within-group benefit on all outcomes at 1-month follow-up and (2) improvement exceeding minimal detectable change, as well as minimal clinically significant difference, on ≥1 outcomes at ≥1 time points after intervention.
Conclusions— After stroke, active peripheral nerve stimulation paired with intensive task–oriented training can effect significant improvement in severely impaired upper extremity movement function. Further confirmatory studies that consider a larger group, as well as longer follow-up, are needed.
Stroke rehabilitation in low- and middle-income countries, such as Mexico, is often hampered by lack of clinical resources and funding. To provide a cost-effective solution for comprehensive post-stroke rehabilitation that can alleviate the need for one-on-one physical or occupational therapy, in lower and upper extremities, we proposed and implemented a technology-assisted rehabilitation gymnasium in Chihuahua, Mexico. The Gymnasium for Robotic Rehabilitation (Robot Gym) consisted of low- and high-tech systems for upper and lower limb rehabilitation. Our hypothesis is that the Robot Gym can provide a cost- and labor-efficient alternative for post-stroke rehabilitation, while being more or as effective as traditional physical and occupational therapy approaches.
A typical group of stroke patients was randomly allocated to an intervention (n = 10) or a control group (n = 10). The intervention group received rehabilitation using the devices in the Robot Gym, whereas the control group (n = 10) received time-matched standard care. All of the study subjects were subjected to 24 two-hour therapy sessions over a period of 6 to 8 weeks. Several clinical assessments tests for upper and lower extremities were used to evaluate motor function pre- and post-intervention. A cost analysis was done to compare the cost effectiveness for both therapies.
No significant differences were observed when comparing the results of the pre-intervention Mini-mental, Brunnstrom Test, and Geriatric Depression Scale Test, showing that both groups were functionally similar prior to the intervention. Although, both training groups were functionally equivalent, they had a significant age difference. The results of all of the upper extremity tests showed an improvement in function in both groups with no statistically significant differences between the groups. The Fugl-Meyer and the 10 Meters Walk lower extremity tests showed greater improvement in the intervention group compared to the control group. On the Time Up and Go Test, no statistically significant differences were observed pre- and post-intervention when comparing the control and the intervention groups. For the 6 Minute Walk Test, both groups presented a statistically significant difference pre- and post-intervention, showing progress in their performance. The robot gym therapy was more cost-effective than the traditional one-to-one therapy used during this study in that it enabled therapist to train up to 1.5 to 6 times more patients for the approximately same cost in the long term.
The results of this study showed that the patients that received therapy using the Robot Gym had enhanced functionality in the upper extremity tests similar to patients in the control group. In the lower extremity tests, the intervention patients showed more improvement than those subjected to traditional therapy. These results support that the Robot Gym can be as effective as traditional therapy for stroke patients, presenting a more cost- and labor-efficient option for countries with scarce clinical resources and funding.
Fig. 1 Distribution of the stations inside facility at CREE Chihuahua. Beginning by the door, clockwise, is: the Theradrive system in the first station; the Ness for upper extremity in the second station; the Ness for lower extremity in the third station; the Motomed Viva 2 for upper extremities in the fourth station; the Motomed Viva 2 for lower extremities in the fifth station; and Capitain’s Log Brain-trainer in the sixth station
Visual field deficits are common in patients with damaged retinogeniculostriate pathways. The patient’s eye movements are often affected leading to inefficient visual search. Systematic eye movement training also called compensatory therapy is needed to allow patients to develop effective coping strategies. There is a lack of evidence-based, clinical gold-standard registered medical device accessible to patients at home or in clinical settings and NeuroEyeCoach (NEC) is developed to address this need. In three experiments, we report on performance of patients on NEC compared to the data obtained previously on the earlier versions of the search task (); we assessed whether the self-administered computerised tasks can be used to monitor the progress () and compared the findings in a subgroup of patients to a healthy control group. Performance on cancellation tasks, simple visual search, and self-reported responses on activities of daily living was compared, before and after training. Patients performed similarly well on NEC as on previous versions of the therapy; the inbuilt functionality for pre- and postevaluation functions was sensitive to allowing assessment of improvements; and improvements in patients were significantly greater than those in a group of healthy adults. In conclusion, NeuroEyeCoach can be used as an effective rehabilitation tool to develop compensatory strategies in patients with visual field deficits after brain injury.
We explore our surrounding environment by moving our eyes on average three times per second. The eye movement episodes are punctuated by brief periods (100–300 ms) of fixations. This pattern of activity ensures detailed image processing by the high density cone-receptor region of our central vision . The resultant continuous perception of the stable world relies on amalgamation of lower resolution peripheral vision with high resolution central information in a spatiotopic frame of reference . This dynamic process encompasses the suppression of noise or distractors and selective enhancement of target objects . The selection of candidate targets for subsequent eye movements (saccades) is achieved through a combination of stimulus driven bottom-up and goal driven top-down mechanisms .
Visual field deficits often accompany lesions of the visual pathways which in turn disrupt the selection of targets falling within the impaired visual fields . Abnormal patterns of eye movement are reported in approximately 60% of such cases . One method for quantifying disturbances of visual processing is to make use of a visual search paradigm where the patient is required to report the presence or absence of a target amongst distractor items, often but not exclusively, presented on a computer screen . The reaction times are then compared to those for target detection in the sighted field in the same individual or in a group of healthy individuals. The inverse of the slope for a linearly fitted plot of reaction times as a function of the number of distractor items reflects “search efficiency” . In general, for healthy adults when targets and distractors are easily discriminable (pop-out search), the slope is shallow (high efficiency), but steeper slopes are expected when targets and distractors share features (complex or conjunction search).
Eye movement recordings of patients with visual field deficits following brain injury reveal a number of characteristics . These include smaller saccade amplitudes, and, hence, a larger number of fixations; limited exploration of the contralesioned visual field; and more between-hemifield saccades often summarised as disorganised eye movements leading to slower reaction times for targets in contralesioned hemifields. Disturbances of eye movement dynamics are also reported in the sighted (ipsilesioned) hemifield [6, 10].
In clinical practice, the rehabilitation of patients with visual field deficits is often conducted by occupational therapists or low-vision experts. The aim of any intervention is to improve the patient’s interactions with their immediate surrounding and increasing their confidence in tasks such as shopping or commuting. The use of computerised visual search tasks as a rehabilitation tool to improve eye movements after brain injury was first reported in a group of 30 patients . Patients were given systematic practice with large saccadic eye movements to search for targets presented at unpredictable positions in both the affected hemifield and the entire field of gaze. This class of treatment was later extended by use of a visual search paradigm to improve scanning strategy. Simultaneous recording of eye movements in a group of 60 patients provided further evidence for spatially disorganised pattern of eye movements in 60% of cases , with improved visual scanning in all 13 cases that underwent visual search training. With better use of the remaining sight as well as efficient search strategy, patients were able to compensate for their partial blindness; hence, the technique has been termed compensatory. This technique with various modifications has been used in 14 studies to date, with a total of 593 patients with homonymous visual field loss and persistent visual disabilities (see Table 1). Indeed a recent systematic review  has identified eye movement training as the most promising approach to visual rehabilitation in stroke patients.
Background and Objective. Favorable prognosis of the upper limb depends on preservation or return of voluntary finger extension (FE) early after stroke. The present study aimed to determine the effects of modified constraint-induced movement therapy (mCIMT) and electromyography-triggered neuromuscular stimulation (EMG-NMS) on upper limb capacity early poststroke.
Methods. A total of 159 ischemic stroke patients were included: 58 patients with a favorable prognosis (>10° of FE) were randomly allocated to 3 weeks of mCIMT or usual care only; 101 patients with an unfavorable prognosis were allocated to 3-week EMG-NMS or usual care only. Both interventions started within 14 days poststroke, lasted up until 5 weeks, focused at preservation or return of FE.
Results. Upper limb capacity was measured with the Action Research Arm Test (ARAT), assessed weekly within the first 5 weeks poststroke and at postassessments at 8, 12, and 26 weeks. Clinically relevant differences in ARAT in favor of mCIMT were found after 5, 8, and 12 weeks poststroke (respectively, 6, 7, and 7 points; P < .05), but not after 26 weeks. We did not find statistically significant differences between mCIMT and usual care on impairment measures, such as the Fugl-Meyer assessment of the arm (FMA-UE). EMG-NMS did not result in significant differences.
Conclusions. Three weeks of early mCIMT is superior to usual care in terms of regaining upper limb capacity in patients with a favorable prognosis; 3 weeks of EMG-NMS in patients with an unfavorable prognosis is not beneficial. Despite meaningful improvements in upper limb capacity, no evidence was found that the time-dependent neurological improvements early poststroke are significantly influenced by either mCIMT or EMG-NMS.
Several prospective cohort studies among stroke patients have shown that the functional outcome of the upper limb is largely defined within the first 5 weeks poststroke and is mainly driven by (yet poorly understood) mechanisms of spontaneous neurological recovery.1,2 Observational studies showed that the presence of some voluntary finger extension (FE) within 72 hours is a favorable indicator for the return of dexterity poststroke.3,4 This suggests that early control of FE is an important prognostic factor in stratifying patients for upper limb intervention trials early poststroke.2
For those with a favorable prognosis, indicated by some voluntary FE early poststroke, constraint-induced movement therapy (CIMT) or a modified version (mCIMT) may benefit arm-hand activities and self-reported hand function in daily life.5The number of phase II trials on mCIMT within the first days or weeks poststroke is however small and findings are rather inconclusive. For example, Dromerick et al6showed in a proof of concept trial that 1 or 2 hours mCIMT per working day for 2 weeks was not superior to an equal dosage of usual care, whereas a high dose of 3 hours mCIMT per working day resulted in less improvement on functional outcome measured with the Action Research Arm Test (ARAT) at 3 months poststroke.
For those with an unfavorable prognosis for functional outcome at 6 months, that is, patients without voluntary FE,1,3,4 no evidence-based therapies have been reported so far. In subacute and chronic stroke, innovative therapies such as electromyography-triggered neuromuscular stimulation (EMG-NMS) of the finger extensors to improve voluntary control have shown promise in terms of increasing active range of motion.7⇓⇓⇓–11 Furthermore, several studies suggest that EMG-NMS may produce changes in cortical activation patterns and excitability in chronic stroke.12,13 For example, Shin et al13 showed in a small proof of concept trial (n = 14) that a daily 30-minute program for 10 weeks shifted cortical activation patterns as seen in functional magnetic resonance imaging from the ipsilateral sensorimotor cortex to the contralateral sensorimotor cortex in chronic stroke. Despite the growing evidence for enhanced levels of homeostatic neuroplasticity in the first weeks poststroke,14 early started EMG-NMS trials for patients without FE are lacking in this restricted time window.
The first objective of the present study was to investigate the effects of an early mCIMT program on recovery of upper limb capacity during the first 6 months, starting within 14 days poststroke in patients with some voluntary FE. Our second objective was to investigate the effects of early EMG-NMS on the recovery of voluntary FE and upper limb capacity during the first 6 months, starting within 14 days poststroke in patients with no voluntary control of the finger extensors. We hypothesized that an intensive 3-week mCIMT program would result in a clinically meaningful improvement in ARAT scores compared with usual care alone. For the patients with an unfavorable prognosis we hypothesized that a higher percentage of patients (10% or more increase) would regain some dexterity (ARAT score >9 points on a maximum of 57 points) if they received intensive daily EMG-NMS for 3 weeks, compared with usual care alone.
Figure 1. Inclusion flow diagram. The total amount of patients with cerebrovascular accidents was estimated using the number of admitted patients in each participating center. mCIMT: modified constrained-induced movement therapy; EMG-NMS, electromyography-triggered neuromuscular stimulation.
The ability to associate neutral stimuli with motivationally relevant outcomes is an important survival strategy. In this study, we used event-related potentials (ERPs) to investigate brain dynamics of associative emotional learning when participants were confronted with multiple heterogeneous information. Participants viewed 144 different objects in the context of 144 different emotional and neutral background scenes. During each trial, neutral objects were shown in isolation and then paired with the background scene. All pairings were presented twice to compare ERPs in response to neutral objects before and after single association. After single pairing, neutral objects previously encoded in the context of emotional scenes evoked a larger P100 over occipital electrodes compared to objects that were previously paired with neutral scenes. Likewise, larger late positive potentials (LPPs) were observed over parieto-occipital electrodes (450–750 ms) for objects previously associated with emotional relative to neutral contexts. The LPP – but not P100 – enhancement was also related to subjective object/context binding. Taken together, our ERP data provide evidence for fast emotional associative learning, as reflected by heightened perceptual and sustained elaborative processing for neutral information previously encountered in emotional contexts. These findings could assist in understanding binding mechanisms in stress and anxiety, as well as in addiction and eating-related disorders.
One important survival strategy is to perceive fluctuating changes that occur in contiguous environments in order to readjust the momentary motivational relevance of incoming information. This ability allows developing flexible and adaptive responses based on the history of contingencies encountered by the individual (Miskovic & Keil, 2012). In this sense, it has been observed that a previously neutral stimulus (conditioned stimulus; CS+) continuously associated with an aversive event (unconditioned stimulus; UCS) acquires motivational relevance, compared to a neutral stimulus (CS−) unpaired with a UCS or associated with a non-emotional UCS, a process called associative learning. Traditionally, a large number of pairings between few and/or simple CS+ and a strongly aversive UCS have been used in learning paradigms to generate strong associations (Lissek et al., 2006). However, rather than single, unambiguous and/or isolated CS/UCS pairings, we are constantly confronted with multiple different events that imply associations between neutral and moderately relevant stimuli. Thus, the use of paradigms involving ‘weak’ ambiguous situations (e.g. less salient UCS, multiple complex pairings and/or few contingencies CS/UCS) would provide a better understanding of the underpinnings of associative learning (Lissek et al., 2006; Beckers et al., 2013; Steinberg et al., 2013b; Hur et al., in press). In the present study, we investigated the role of UCS heterogeneity on the formation of associations using electrophysiological correlates of associative learning for multiple neutral events paired with multiple emotional contingencies (emotional scenes).
Recent studies from Junghöfer and colleagues (e.g. Pastor et al., 2015; see Steinberg et al., 2013b, for review) used the so-called MultiCS conditioning, in which multiple CSs+ (e.g. pictures of different faces) were associated with emotionally relevant UCSs (e.g. aversive and appetitive sounds, electric shocks), while other CSs− remained unpaired or were associated with neutral events. Brain activation was measured using electro- and magnetoencephalography (EEG, MEG) during these conditioning procedures. After multiple pairings, CSs+ compared to CSs− evoked enhanced neural activity at prefrontal and sensory cortical regions during earlier (< 300 ms; Bröckelmann et al., 2011; Steinberg et al., 2012, 2013a; Rehbein et al., 2014, 2015) and later stages of processing (> 300 ms; Pastor et al., 2015), irrespective of contingency awareness. These results suggest the existence of a rather automatic learning mechanism that rapidly transfers the emotional properties of the UCS to CSs, leading to a facilitated perceptual and a more elaborated processing of the CS+.
Nevertheless, these studies have only used highly salient UCS. Therefore, it is unclear whether such associative learning processes also occur in the presence of less intense emotional events – i.e., reproducing daily interactions – or whether the formation of associations is exclusively facilitated in survival-specific contexts (Öhman & Mineka, 2001). It is also unclear whether the acquired motivational significance leading to neural response enhancement for emotion-associated stimuli occurs rapidly after one single pairing (e.g. Morel et al., 2012; Rehbein et al., 2014), or whether more than one repetition is needed to form such associations (e.g. Steinberg et al., 2012). While most of the electrophysiological conditioning studies have used aversive cues as UCS (see Miskovic & Keil, 2012, for review), it has recently been observed that pleasant information can also serve as effective, intrinsically motivating UCSs (Schacht et al., 2012; Steinberg et al., 2013a; Blechert et al., 2016; see Martin-Soelch et al., 2007, for neuroimaging findings). Both aversive and appetitive conditioning processes likely not only contribute to various disorders, such as trauma- and stress-related disorders, but also to substance abuse and eating-related disorders (e.g. Martin-Soelch et al., 2007; Pape & Pare, 2010). Thus, more evidence regarding the effect of valence on associative conditioning is needed.
In the present study, we therefore investigated brain dynamics of associative emotional learning when participants viewed neutral objects in the context of different emotionally arousing (both pleasant and unpleasant) and neutral background scenes. Object and scene presentation occurred always in the same order; first objects were presented in isolation (CS) and then a picture scene was added as background (see Fig. 1). Pairings were presented in two consecutive blocks, allowing to compare the processing of CS+ objects – paired with emotional scenes – and CS− objects – paired with neutral scenes, before (first block) and after single pairing (second block). Based on previous EEG and MEG conditioning studies (see Miskovic & Keil, 2012, for review), we predicted enhanced processing of neutral cues previously paired with emotional contexts, irrespective of valence, relative to cues previously paired with neutral contexts at different stages of processing. Because both perceptual and sustained elaborative processing have been found to be enhanced for stimuli associated with CS+, we predicted enhanced positivity for the CS+ compared with the CS− at (a) earlier (P100) and (b) later stages of processing [late positive potential (LPP)].
Figure 1. Schematic view of the stimulus presentation during the first and the second associative learning blocks. When an object was seen during the first block, the object/background association has not yet taken place, and when the object was seen in the second block, object and background scene have been associated once. [Correction added on 19 Aug 2016, after original online publication: Figure 1 has been corrected.]
Rhythmic and discrete upper-limb movements are two fundamental motor primitives controlled by different neural pathways, at least partially. After stroke, both primitives can be impaired. Both conventional and robot-assisted therapies mainly train discrete functional movements like reaching and grasping. However, if the movements form two distinct neural and functional primitives, both should be trained to recover the complete motor repertoire. Recent studies show that rhythmic movements tend to be less impaired than discrete ones, so combining both movement types in therapy could support the execution of movements with a higher degree of impairment by movements that are performed more stably.
A new performance-based assistance method was developed to train rhythmic movements with a rehabilitation robot. The algorithm uses the assist-as-needed paradigm by independently assessing and assisting movement features of smoothness, velocity, and amplitude. The method relies on different building blocks: (i) an adaptive oscillator captures the main movement harmonic in state variables, (ii) custom metrics measure the movement performance regarding the three features, and (iii) adaptive forces assist the patient. The patient is encouraged to improve performance regarding these three features with assistance forces computed in parallel to each other. The method was tested with simulated jerky signals and a pilot experiment with two stroke patients, who were instructed to make circular movements with an end-effector robot with assistance during half of the trials.
Simulation data reveal sensitivity of the metrics for assessing the features while limiting interference between them. The assistance’s effectiveness with stroke patients is established since it (i) adapts to the patient’s real-time performance, (ii) improves patient motor performance, and (iii) does not lead the patient to slack. The smoothness assistance was by far the most used by both patients, while it provided no active mechanical work to the patient on average.
Our performance-based assistance method for training rhythmic movements is a viable candidate to complement robot-assisted upper-limb therapies for training a larger motor repertoire.
Rhythmic and discrete movements have recently been recognized as two of the most fundamental units of the upper-  and lower-limb  motor repertoire. Rhythmic movements capture periodic movements like hammering or scratching, while discrete movements capture movements between a succession of postures with zero velocity and acceleration, like reaching and pointing [3, 4]. These two fundamental motor primitives are controlled by distinct neural circuitries, at least partially [3, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14]. For example, previous research with healthy subjects showed that (i) discrete movements require more cortical activity than rhythmic ones , and (ii) no learning transfer occurs from rhythmic to discrete movements and only a little transfer occurs from discrete to rhythmic movements when they are executed in altered visual or haptic conditions [13, 14].
After a stroke, both rhythmic and discrete movements can be impaired [15, 16, 17, 18, 19, 20, 21, 22, 23]. Recently, we compared the performance in executing both movements in the same stroke population. As a main conclusion, we found that rhythmic arm movements are less affected than discrete ones. In particular, stroke preserved the smoothness of rhythmic movements so that fewer submovements were identified than in the discrete counterparts . However, rhythmic movements were impaired compared to healthy subjects. Stroke patients decelerated more than healthy subjects at the movement reversal, and some patients displayed a larger amount of submovements.
If rhythmic and discrete movements are two distinct primitives, they deserve specific and differentiated training to permit the full recovery of the complete motor repertoire. This is a necessary condition to recover autonomy life activities requiring a combination of rhythmic and discrete movements (such as wiping a table or playing the piano ).
Most post-stroke therapies tend to focus on functional and thus mainly discrete movements [25, 26, 27], although some previous contributions did focus on upper-limb rhythmic movement training. Interestingly, they all tended to display an improvement in motor skills [28, 29, 30, 31, 32, 33, 34]. For instance, [28, 29] highlighted that the intensity of the training is critical to enhance motor skills. In , the authors compared bilateral arm training with auditory cueing (BATRAC) to dose-matched therapeutic exercises and concluded that none was superior to the other, although the adaptations in brain activation were greater after BATRAC. Whether this result is due to the rhythmic nature of the movement, its bimanual nature, the auditory cueing, or a combination of these features, is however difficult to establish, since these are closely intertwined in BATRAC.
The current state-of-art of rhythmic upper-limb movement therapy calls thus for the development of post-stroke therapies tailored to unilateral rhythmic movement training, in order to study their exact effect on motor skills. The development of such a therapy is presented in this paper.
Robotic devices are particularly suited for implementing post-stroke therapies, with a specific focus on movement intensity. Rehabilitation robots enable patients to practice well-specified motor actions and can deliver an appropriate amount of assistance to help patients in improving their motor behavior [17, 35, 36, 37, 38, 39, 40, 41, 42]. Motor performance can be computed in real-time by the robot controller, allowing for continuous adaptation of the type and amount of assistance. The patient only receives the necessary support and is prevented from slacking [36, 43]. In the literature, this is often referred to as the “slacking hypothesis,” which suggests that too much assistance will cause a progressive decrease in patient effort to accomplish a desired task and reduce motor recovery. This assistance principle is also called “assistance as needed” and has progressively emerged as a hallmark of successful robot-assisted therapies [35, 36, 43, 44]. This principle lies also at the core of the present contribution.
Most upper-limb robot-assisted therapies are designed for discrete movement training and implement the assist-as-needed principle through different strategies. One type of strategy delivers assistance proportional to the trajectory error with respect to a predefined trajectory [42, 45, 46, 47, 48, 49, 50]. Another assistance approach relies on dynamical systems and adapts the assistance parameters as a function of the patient performance . Other approaches tune the amount of assistance across sessions as a function of the performance during the preceding session [45, 51]. Another method  performs an online adaptation of the amount of support depending on the activity (for a survey, see ).
In contrast with these approaches, a rhythmic movement therapy should exploit the cyclic nature of the movement to anticipate the future trajectory based on previous cycles. This can be achieved by using adaptive oscillators . These mathematical tools are particularly suited to track the main features of a typical rhythmic movement (like amplitude and frequency). This continuous assessment allows the robot to constantly seek to improve movement features with the appropriate amount and type of assistance. Moreover, this approach naturally allows for trajectory-free assistance algorithms so that the therapist does not have to specify an arbitrary target trajectory for the patient to follow. The patient receives assistance to improve the impacted movement features, but is left free to produce any rhythmic trajectory.
Our previous work already paved the way in using adaptive oscillators to deliver trajectory-free assistance for upper-  and lower-limb [54, 55] rhythmic movements. These contributions focused on movement assistance for healthy subjects, showing evidence of decreases in metabolic consumption when the assistance was switched on. The present study is the first to propose a metric-based assistance method for patients with motor disorders, with emphasis on the potential to assist different rhythmic movement features as a function of the patient needs.
This paper outlines the performance-based assistance method and its mathematical foundations in details. The method was validated with data from simulations and a pilot study with two stroke patients with upper-limb impairments is also reported. The proposed performance-based assistance method can (i) enhance motor-performance, (ii) give appropriate assistance according to patient performance, and (iii) maintain active patient participation in the task so that no slacking effect occurs.
The main interest of the developed method is that it can independently assist different movement features, only if needed. In particular, the method implements parallel strategies to assist the patient in improving performance regarding movement smoothness, velocity, and amplitude. Therefore, the method requires measuring (Fig. 1a) and quantifying (Fig. 1b) the amplitude, velocity, and smoothness features of patient movement in real-time in order to assess the corresponding performance. The method must also compute and deliver the appropriate amount of assistance in amplitude, velocity, and/or smoothness as a function of the performance (Fig. 1c).
Fig. 1 Methodology. Outline of the overall control strategy of the performance-based assistance. First, the movement features are computed by the adaptive oscillator (a) and serve as input to compute the real-time performance in smoothness, velocity, and amplitude (b). These features are then used to compute the gains to tune the level of the assistance forces in smoothness, velocity, and amplitude (c). These three forces are eventually summed up and delivered to the patient