Medical Policy

This document addresses brain computer interface (BCI) rehabilitation devices. These are non-invasive devices designed to enable individuals, particularly those with neurological conditions such as stroke, to control external devices using their brain signals.

Note: For further information on similar technologies, please see the following related document:

• OR-PR.00005 Upper Extremity Myoelectric Orthoses

Note: For a high-level overview of this document, please see “Summary for Members and Families” below. 

Investigational and Not Medically Necessary:

Brain computer interface rehabilitation devices, including but not limited to electroencephalography (EEG)-driven upper extremity powered exercisers, are considered investigational and not medically necessary.

This document describes clinical studies and expert recommendations, and explains why we do not consider the use of devices that use brain wave signals (EEGs) to help move or exercise a body part to be clinically appropriate. Such devices are called brain computer interface devices. The following summary does not replace the medical necessity criteria or other information in this document. The summary may not contain all relevant criteria or information. This summary is not medical advice. Please check with your healthcare provider for any advice about your health.

Key Information

Available brain computer interface (BCI) rehabilitation devices are designed to help people move a weak or paralyzed arm or hand by using signals from the brain. One example is the IpsiHand™ Upper Extremity Rehabilitation System (IpsiHand) (Neurolutions, Inc., Saint Louis, MO). The device uses brain signals to control a hand exerciser during stroke rehabilitation. Possible benefits include improved hand and arm movement after a stroke. However, the device may not work for everyone, and studies have not yet shown that it improves long-term health outcomes. Better studies are needed to know if this treatment improves health.

What the Studies Show

The IpsiHand uses electroencephalography (EEG), a method that detects electrical activity in the brain through sensors placed on the scalp, and sends the signals to a hand device that helps open and close the affected hand during rehabilitation exercises. The goal is to help people regain movement after a stroke by strengthening the nerve connections in the brain and from the brain to the hand.

Two published studies evaluated the IpsiHand. In a 2022 study, 17 people who had experienced a stroke used the device for 12 weeks. The study found changes in brain activity that were linked to improved movement during therapy. However, the study did not include a comparison group, so it is unclear whether the improvements would have happened even without the device. In a 2025 study, some participants showed improved arm and hand movement after treatment. However, the study also lacked a comparison group, many participants did not complete the study, and the researchers did not evaluate long-term results. In addition, some people could not produce the brain signals needed for the device to work. These problems make it difficult to know whether the device improves health or whether the results would be similar in a larger group of people.

Is this clinically appropriate?

Brain computer interface rehabilitation devices, including electroencephalography (EEG)-driven upper extremity powered exercisers such as the IpsiHand, are not clinically appropriate because they have not been proven to improve health.

The available studies reported some improvements in movement and brain activity. However, the studies had important problems that limit the ability to apply the results to other people. Unnecessary or unproven treatments can lead to unnecessary risks and delay in getting helpful treatment.

(Return to Description/Scope)

Summary

The evidence regarding noninvasive BCI rehabilitation after stroke has evolved and now includes randomized comparative studies and several systematic reviews and meta-analyses. In a meta-analysis of 32 randomized trials involving 1,187 participants, Mortezaei and colleagues (2026) found that BCI-assisted rehabilitation produced greater improvement in upper-extremity Fugl-Meyer scores than control rehabilitation (mean difference, 3.85 points; 95% CI, 2.84-4.86). However, significant between-group improvements were not demonstrated for the Action Research Arm Test, Motor Activity Log, Modified Barthel Index, or Modified Ashworth Scale. The included studies evaluated heterogeneous BCI technologies and co-interventions, limiting applicability to any individual commercial device.

Product-specific evidence for IpsiHand now includes a randomized controlled trial. Leuthardt and colleagues (2026) reported greater improvement in UEFM scores with IpsiHand than with a structured home exercise program after 12 weeks of treatment. The prespecified primary analysis estimated a between-group difference of 4.5 points, and clinically meaningful improvement was estimated in 55.5% of BCI participants compared with 9.6% of controls. However, a treatment-policy sensitivity analysis produced a smaller between-group difference of 2.5 points. Interpretation is limited by substantial differential post-randomization attrition related to treatment assignment, absence of a sham control and participant blinding, selected enrollment requiring an adequate EEG control signal, lack of post-treatment follow-up, and absence of significant differences for most secondary functional outcomes.

Prasad and colleagues (2026) subsequently reported retrospective real-world outcomes among individuals prescribed IpsiHand. Among 56 analyzed individuals, 64% met the study's 5.25-point UEFM improvement threshold within 12 weeks and 70% met the threshold at some point during variably available follow-up extending to 55 weeks. However, 22 of 78 potentially eligible individuals were excluded because missing assessments prevented response classification, follow-up became increasingly sparse over time, treatment adherence was uncertain, and there was no comparator. The study therefore provides supportive evidence that motor improvement may occur during continued IpsiHand use but does not establish the magnitude of benefit attributable to the device, durability after treatment cessation, or the effectiveness of extended treatment among individuals who have not responded by 12 weeks.

Overall, the available evidence supports a treatment effect of noninvasive BCI rehabilitation, including IpsiHand, on measures of upper-extremity motor impairment after stroke. However, improvement has not been consistently demonstrated in activity-level outcomes, real-world arm use, independence in activities of daily living, or other broader functional outcomes, and durable benefit after treatment cessation has not been established. Product-specific evidence remains dependent primarily on a single randomized trial with important methodological limitations and has not been independently replicated. Therefore, the available evidence does not yet establish IpsiHand as clinically appropriate and effective in accordance with generally accepted standards of medical practice.

Discussion

To date, only one BCI device has been successfully reviewed by the Food and Drug Administration (FDA), the IpsiHand Upper Extremity Rehabilitation System.

A preliminary study examined changes in the coordination of brain rhythms during use of the IpsiHand system (Rustamov, 2022). The study included 17 individuals who had a stroke at least 6 months prior to study participation. Other eligibility criteria included intact cognitive ability, unilateral upper extremity weakness and normal sensation in the affected upper extremity. Participants used the IpsiHand system for 12 weeks as part of a rehabilitation therapy program. Brain signal measurements were taken at the beginning of therapy and every 4 weeks during therapy. The investigators found that the coordination between theta and gamma brain rhythms improved in both left and right brain regions controlling movement, and the improvement in brain rhythms were associated with improvements seen during therapy sessions. The study was not designed to evaluate whether use of the IpsiHand system improved clinical outcomes. It did not include a control group of individuals who did not use the IpsiHand system and therefore it is unclear whether changes in motor functioning were due to the use of the device or some other factor such as added attention given to trial participants.

Ji (2025) conducted a prospective, single-center, randomized controlled trial to evaluate the effectiveness and neural mechanisms of a BCI controlled soft robotic glove (BCI-SRG) for upper-limb rehabilitation in individuals with subacute stroke. A total of 40 individuals with left-sided hemiparesis occurring 2 weeks to 3 months after a right hemispheric stroke were randomly assigned to either a BCI-SRG group (n=20) or a soft robotic glove (SRG) group (n=20). Both groups received conventional upper-limb rehabilitation along with 20 intervention sessions over 4 weeks. The BCI-SRG group underwent EEG-based motor imagery training that controlled a soft robotic glove, while the SRG group used the same glove without BCI control. Outcomes were assessed using the Action Research Arm Test (ARAT), Fugl-Meyer Assessment for Upper Limb (FMA-UL), Modified Barthel Index (MBI), and functional near-infrared spectroscopy (fNIRS). Thirty-nine participants completed the study. Although both groups demonstrated significant functional improvements, the BCI-SRG group achieved greater gains in ARAT scores (median change: 14 points compared with 7 points; p=0.032) and FMA-UL scores (median change: 18 points compared with 7 points; p=0.010). No significant between-group difference was observed for MBI scores (p=0.065). fNIRS findings showed significant increases in cortical activation within the bilateral sensorimotor cortex and medial prefrontal cortex in the BCI-SRG group, whereas no significant activation changes were observed in the SRG group. In addition, increased prefrontal cortical activation was positively correlated with improvements in ARAT scores, suggesting a relationship between cognitive-motor network engagement and upper-limb functional recovery. The findings suggest that BCI-assisted robotic rehabilitation may enhance upper-limb motor recovery compared with robotic glove therapy alone; however, interpretation is limited by the small sample size, short duration of follow-up, and single-center design.

Kim (2025) conducted a double-blinded, parallel-group, randomized controlled trial to investigate the effects of BCI training with motor imagery (MI)-contingent feedback on upper-limb function and neuroplasticity in individuals with chronic stroke. Twenty-seven participants with chronic stroke (≥6 months post-stroke) and severe wrist extensor weakness were randomly assigned to either an MI-contingent feedback BCI group (n=14) or an MI-independent feedback BCI group (n=13). Both groups completed 20 sessions of BCI training over 4 weeks, consisting of EEG-based motor imagery tasks combined with functional electrical stimulation (FES) of the affected wrist extensors. The MI-contingent feedback group received FES only when the BCI system detected correct motor imagery, whereas the MI-independent feedback group received FES regardless of motor imagery performance. Outcome measures included wrist extensor muscle strength measured using the Medical Research Council scale (MRC-WE), active range of motion in wrist extension (AROM-WE), FMA, Box and Block Test (BBT), Stroke Impact Scale (SIS), and resting-state electroencephalography to evaluate functional connectivity. Twenty-five participants completed the study. Participants receiving MI-contingent feedback demonstrated significantly greater improvements in MRC-WE scores at 4 weeks (mean difference [MD], 0.52; 95% Confidence Interval [CI], 0.03-1.00; p=0.036) and showed significant increases in AROM-WE (p=0.019). Significant within-group improvements in AROM-WE were observed only in the MI-contingent feedback group, although the between-group difference was not statistically significant. EEG analyses demonstrated changes in functional connectivity within motor-related brain regions that correlated with improvements in wrist strength and distal FMA score; however, no significant between-group differences were observed for FMA, BBT, SIS, or spasticity outcomes. The findings suggest that MI-contingent feedback may enhance wrist-specific motor recovery and associated neurophysiologic measures, although broader functional benefits were not demonstrated.

Liu (2025) conducted an umbrella review of systematic reviews and meta-analyses to evaluate the efficacy and safety of BCI interventions for stroke rehabilitation. A comprehensive literature search of major English and Chinese databases through October 2024 identified 18 eligible systematic reviews and meta-analyses investigating BCI-based rehabilitation in individuals with stroke. The methodological quality of the included reviews was assessed using AMSTAR-2 and PRISMA criteria, along with evaluations of study design, homogeneity, publication year, and publication bias. The included reviews examined a range of BCI interventions, either alone or combined with therapies such as conventional rehabilitation, motor imagery training, robotic assistance, functional electrical stimulation, or transcranial direct current stimulation. Across the included evidence, BCI-based rehabilitation consistently demonstrated beneficial effects on upper-limb motor recovery, with improvements reported in outcomes such as the FMA, ARAT, muscle strength, and activities of daily living. Evidence also suggested that BCI interventions may have effects on neuroplasticity and brain function recovery, with the greatest benefits observed in individuals with subacute stroke. In addition, BCI combined with other rehabilitation approaches was associated with improvements in quality of daily life and demonstrated a favorable safety profile. However, the findings should be interpreted cautiously because the included reviews were characterized by substantial methodological limitations; AMSTAR-2 assessments rated 12 reviews as critically low quality and the remaining 6 as low quality. The evidence supporting improvements in speech function, lower-limb motor recovery, and long-term outcomes was further limited by heterogeneity in BCI technologies, co-interventions, outcome measures, and study populations. While BCI interventions appeared to have an acceptable safety profile, adverse-event reporting was limited. Evidence supporting improvements in speech function, lower-extremity motor recovery, and long-term outcomes remains insufficient, and additional high-quality randomized trials and systematic reviews are needed to clarify the durability and clinical significance of observed benefits.

In 2025, Rustamov and colleagues published findings of a single-arm study evaluating IpsiHand. Participants were eligible if they had experienced a stroke at least 6 months prior and had unilateral upper extremity weakness. Initially, 100 individuals agreed to participate, but 31 did not meet all eligibility criteria, and 15 could not consistently produce identifiable BCI control signals during baseline screening. This left 54 participants who began the intervention. Out of those, 24 participants (44%) dropped out, and 30 individuals completed the 12-week treatment program. Motor function was assessed prior to treatment and at 12 weeks by physical and occupational therapists. Participants were then trained how to use the BCI device. Participants were instructed to use the device for 1 hour a day, 5 days a week, for 12 weeks. At the end of the trial, the participants were reassessed by the same therapist who conducted their initial evaluation. The primary tool used to assess motor outcomes was the upper extremity Fugl-Meyer tool (UEFM). The mean increase in the UEFM score was 8.1 points which exceeded the minimal clinically significant difference (MCID) threshold of 5.25 points. Of the 26 individuals with available data, 18 (69%) showed an improvement of at least 5.25 points. The change in UEFM score was significantly correlated with reported usage of the BCI device during the trial. This trial lacked a control group, follow-up beyond the 12- week intervention period, and had a very high dropout rate. Additionally, 15 of the 69 individuals screened for participation (22%) could not produce identifiable BCI control signals, raising concern about the potential applicability and generalizability of the study’s results.

Chen and Yun (2026) conducted a systematic review and meta-analysis of randomized controlled trials to evaluate the efficacy of BCI therapy for upper-limb rehabilitation in individuals with chronic stroke and to identify optimal intervention parameters. A comprehensive search of 6 databases through October 2025 identified 21 randomized controlled trials involving 650 participants with chronic stroke. The primary outcomes included upper-limb motor function assessed using the Fugl-Meyer Assessment for Upper Extremity (FMA-UE) and ARAT, muscle tone assessed using the Modified Ashworth Scale (MAS), and activities of daily living evaluated using the MBI and Motor Activity Log (MAL). Meta-analysis demonstrated that BCI training significantly improved upper-limb motor function (FMA-UE: MD, 2.50, 95% CI, 0.60-4.40; p=0.01) and activities of daily living, as reflected by improvements in MBI (MD, 8.38, 95% CI, 2.23-14.53; p=0.02) and MAL scores (MD, 2.09, 95% CI, 0.42-3.76; p=0.03). However, no significant effects were observed for fine motor function measured by ARAT or muscle tone measured by MAS. The authors noted that the magnitude of improvement in FMA-UE scores was modest and that prediction intervals indicated substantial variability in treatment effects across clinical settings. Subgroup analyses suggested greater benefit with BCI combined with functional electrical stimulation (BCI-FES) and with shorter, higher-frequency treatment protocols; however, formal tests for subgroup differences were not statistically significant, limiting confidence in these findings. Follow-up analyses did not demonstrate sustained benefits after treatment cessation. Overall, the evidence was rated as low to moderate certainty due to methodological limitations and risk of bias in many of the included studies.

In 2026, Mortezaei and colleagues published a systematic review and meta-analysis of 32 randomized trials involving 1,187 individuals evaluating noninvasive brain-computer interface rehabilitation after stroke. The included studies evaluated heterogeneous BCI approaches, including motor imagery-based BCI combined with orthoses, exoskeletons or robotic devices, functional or neuromuscular electrical stimulation, virtual reality or perceptual feedback, neurofeedback, and conventional rehabilitation. Across 27 studies, BCI rehabilitation produced greater improvement in upper-extremity Fugl-Meyer scores than control rehabilitation (mean difference, 3.85 points; 95% CI, 2.84-4.86). However, significant between-group improvements were not demonstrated for the Action Research Arm Test, Motor Activity Log, Modified Barthel Index, or Modified Ashworth Scale. Interpretation is limited by substantial clinical heterogeneity in BCI technologies, co-interventions, stroke populations, and comparators. The authors note the presence of moderate statistical heterogeneity and risk-of-bias concerns in most included trials. The findings support an effect of BCI rehabilitation on upper-extremity motor impairment, but do not establish consistent improvement in activity-level or daily functional outcomes or the effectiveness of any specific commercial BCI device.

Leuthardt and colleagues (2026) conducted a randomized controlled trial to compare an at-home BCI therapy system (IpsiHand) with a structured home exercise program for improving upper-extremity function in adults 18 to 85 years of age with upper-extremity paresis or plegia at least 6 months after stroke. Participants were an average of approximately 5.5 years post-stroke. Before randomization, participants were required to demonstrate an EEG signal sufficient to control the device. Eligible participants were randomized to 12 weeks of at-home BCI therapy using an EEG-based system coupled to a range-of-motion assist handpiece or to a structured home exercise rehabilitation program. Both groups were instructed to complete therapy for 1 hour per day, 5 days per week. Of 109 individuals assessed for eligibility, 85 were randomized, including 43 to the BCI group and 42 to the control group. Seventeen control participants declined further participation because of dissatisfaction with their assigned study group, resulting in substantial differential post-randomization attrition; 62 participants were analyzed for the primary outcome, including 37 in the BCI group and 25 in the control group. The primary outcome was change in the UEFM score from baseline to 12 weeks, with a clinically meaningful response defined prospectively as an improvement of at least 5.25 points. Using the prespecified primary estimand, which incorporated multiple imputation and a hypothetical strategy intended to estimate benefit when the device was used as planned, estimated UEFM improvement was greater with BCI than with home exercise (6.0 points; 95% CI, 3.9-8.1; p<0.0001 versus 1.5 points; 95% CI, −0.1 to 3.1; p=0.07), with an estimated treatment difference of 4.5 points (95% CI, 1.9-7.1; p=0.0007). The estimated clinically meaningful response rate was 55.5% with BCI compared with 9.6% with home exercise. These results correspond to an absolute difference of 45.8% (95% CI, 20.9%-70.8%; p=0.0003) and a number needed to treat of 2.2. A treatment-policy sensitivity analysis that accounted more conservatively for missing outcomes reported a smaller between-group difference of 2.5 points (95% CI, 0.3-4.6; p=0.024). Most prespecified secondary outcomes, including spasticity, gross grasp strength, modified Rankin Scale, and other measures of function, did not differ significantly between groups. Although UEFM assessments were performed by unblinded clinicians, blinded review of recorded assessments found essentially no bias in change scores. No serious or other adverse events were reported. The trial was stopped early for efficacy following a prespecified interim analysis. Interpretation is limited by substantial differential attrition related to treatment assignment, absence of a sham control and participant blinding, selected enrollment requiring an adequate EEG control signal, short-term assessment without post-treatment follow-up, limited power for secondary outcomes, and lack of independent product-specific replication. The findings provide randomized evidence that IpsiHand can improve upper-extremity motor impairment compared with structured home exercise in selected individuals with chronic stroke, but uncertainty remains regarding the magnitude of benefit under routine use and whether improvement in UEFM translates into sustained improvements in broader functional outcomes.

Prasad and colleagues (2026) conducted a retrospective real-world evidence study evaluating upper-extremity motor outcomes in individuals with chronic stroke who were prescribed the IpsiHand BCI system as part of routine care. The investigators analyzed deidentified data collected by the manufacturer’s team for quality-assurance purposes. Eligible individuals were 18 to 85 years of age, had experienced a stroke at least 6 months earlier, had hemiparesis or hemiplegia, and had baseline and follow-up UEFM assessments obtained within 1 week of beginning treatment and at least one follow-up assessment. Participants were instructed to use the at-home BCI device for 1 hour per day, 5 days per week. The study categorized participants as early responders if they achieved the minimal clinically important difference (MCID) of at least 5.25 UEFM points by 6 weeks, intermediate responders if they achieved the MCID by 12 weeks but not by 6 weeks, or early nonresponders if they did not achieve the MCID by 12 weeks. Because missing 6- or 12-week assessments prevented response classification, 22 of 78 individuals (28%) were excluded, leaving 56 participants in the analysis. The mean age was 61.7 years, the mean baseline UEFM score was 18.9, and 46 of 56 participants (82.1%) had severe upper-extremity motor impairment at baseline. Among the 56 analyzed participants, 29 (52%) achieved the MCID by 6 weeks, and an additional 7 met the threshold by 12 weeks, resulting in 36 participants (64%) meeting the threshold within 12 weeks. During variably available follow-up extending to 55 weeks, 39 participants (70%) had at least one assessment meeting the threshold. Among the 46 participants with severe baseline impairment, 29 met the threshold by 12 weeks and 15 were observed to transition to a less severe UEFM impairment category during follow-up. Longer-term data were limited. Only 7 of 20 early nonresponders had UEFM assessments between weeks 29 and 55, of whom 3 subsequently met the 5.25-point threshold. Because participants generally continued using the device, these observations reflect improvement during prolonged treatment rather than durability of benefit after treatment cessation. Interpretation is limited by the retrospective uncontrolled design, irregular and missing outcome assessments, exclusion of 28% of potentially eligible individuals because of missing data, uncertain treatment adherence and concomitant rehabilitation, and absence of functional outcomes beyond the UEFM impairment measure. In addition, the 5.25-point threshold used to define clinically meaningful improvement was derived primarily from a population with less severe impairment than the predominantly severely impaired population in this study, limiting certainty regarding its clinical significance. The longer-term findings are based on small and selectively followed subsets and do not establish the optimal duration of treatment or whether extending IpsiHand therapy beyond 12 weeks benefits individuals who have not responded by that time. Overall, the findings provide supportive evidence that motor improvement can occur during real-world IpsiHand use, including among individuals with severe chronic impairment, but the absence of a comparator and substantial missing data preclude determining the magnitude of treatment effect attributable to IpsiHand or establishing long-term effectiveness.

Current U.S. clinical guidance has not established noninvasive brain-computer interface rehabilitation as recommended standard treatment following stroke. The 2024 VA/DoD Clinical Practice Guideline found insufficient evidence to recommend for or against noninvasive BCI for improving upper-extremity motor outcomes or activities of daily living. However, that guideline's evidence review extended only through May 2023 and therefore predates recent randomized and meta-analytic evidence demonstrating improvement in upper-extremity motor impairment. (VA/DoD, 2024) More recent AHA/ASA guidance emphasizes access to evidence-based stroke rehabilitation and the importance of patient-centered and long-term functional outcomes but does not provide a specific recommendation regarding BCI rehabilitation. IpsiHand is FDA-authorized for selected adults with chronic stroke to facilitate muscle re-education and maintain or increase upper-extremity range of motion; FDA authorization does not establish that the device improves broader functional outcomes or represents generally accepted rehabilitation practice. (FDA, 2026)

BCI rehabilitation devices are technologies that detect brain signals related to the intent for movement and can then use these signals to control a limb, computer or digital device. Many types of BCI systems are in development and/or undergoing clinical trials. One of the primary potential applications of BCI technology is as an aid for motor recovery in individuals who have had a stroke, such as by translating brain signals into movement of paralyzed limbs or into the control of orthotic devices.

At least one BCI device is commercially available. It is used as an aid in post-stroke upper limb rehabilitation. This device, the IpsiHand Upper Extremity Rehabilitation System (Neurolutions, Inc.) went through the FDA de Novo classification process and, in April 2021 was classified as a type II device with the generic name of an EEG driven upper extremity powered exerciser (DEN200046). The device type was defined as a “non-invasive prescription device intended for rehabilitation by driving movement or exercise of an impaired upper extremity in response to the detection of purpose oriented electrical activity produced by the patient's brain.” The FDA de Novo classification process did not require that clinical data be submitted demonstrating that the device improves the net health outcome.

According to the FDA:

The Neurolutions IpsiHand Upper Extremity Rehabilitation System is indicated for use in chronic stroke patients (≥ 6 months post-stroke) age 18 or older undergoing stroke rehabilitation, to facilitate muscle re-education and for maintaining or increasing range of motion in the upper extremity.

The IpsiHand system consists of three components, a headset, a tablet and a handpiece. The non-invasive EEG headset is designed to measure electrical signals from the uninjured side of the brain regarding the user’s intention to move the affected hand. The EEG data is transmitted wirelessly to the tablet, which serves as an interface between the headset and handpiece. A signal is sent to the wireless handpiece via the tablet, and the handpiece then opens or closes the user’s hand. The system may be used as part of a rehabilitation program to help the user improve grasping, with the tablet guiding therapy sessions.

Brain computer interface (BCI): A system that detects and interprets neural activity and translates that activity into commands that control an external device or provide feedback to the user.

Chronic stroke: For purposes of this document, stroke occurring at least 6 months previously.

EEG-driven upper extremity powered exerciser: A noninvasive prescription rehabilitation device that detects purpose-oriented electrical activity from the brain and uses that activity to drive movement or exercise of an impaired upper extremity.

Electroencephalography (EEG): A noninvasive method of recording electrical activity in the brain using electrodes placed on the scalp.

Functional electrical stimulation (FES): Electrical stimulation applied to peripheral nerves or muscles to produce or assist a functional movement.

Minimal clinically important difference (MCID): The smallest change in an outcome measure considered likely to represent a meaningful change in an individual's clinical status. The MCID may vary according to the population, baseline impairment, outcome measure, and method used to establish the threshold.

Motor imagery: The mental rehearsal or imagined performance of a movement without actually performing the movement.

Noninvasive brain-computer interface: A BCI that detects brain activity without surgically implanted electrodes or other intracranial components. Noninvasive BCIs commonly use scalp EEG to detect neural signals.

Upper Extremity Fugl-Meyer Assessment (UEFM or FMA-UE): A standardized assessment of motor impairment after stroke that evaluates movement, coordination, and related motor function of the upper extremity; higher scores indicate better motor performance.

The following codes for treatments and procedures applicable to this document are included below for informational purposes. Inclusion or exclusion of a procedure, diagnosis or device code(s) does not constitute or imply member coverage or provider reimbursement policy. Please refer to the member's contract benefits in effect at the time of service to determine coverage or non-coverage of these services as it applies to an individual member.

When services are Investigational and Not Medically Necessary:
For the following procedure code, or when the code describes a procedure indicated in the Position Statement section as investigational and not medically necessary.

Peer Reviewed Publications:

1. Chen H, Yun G. Efficacy of brain-computer interface therapy for upper limb rehabilitation in chronic stroke: systematic review and meta-analysis of randomized controlled trials. J Med Internet Res. 2026; 28:e79132.
2. Ji X, Lu X, Xu Y, et al. Effects and neural mechanisms of a brain-computer interface-controlled soft robotic glove on upper limb function in patients with subacute stroke: a randomized controlled fNIRS study. J Neuroeng Rehabil. 2025; 22(1):171.
3. Kim MS, Park H, Kwon I, et al. Efficacy of brain-computer interface training with motor imagery-contingent feedback in improving upper limb function and neuroplasticity among persons with chronic stroke: a double-blinded, parallel-group, randomized controlled trial. J Neuroeng Rehabil. 2025; 22(1):1.
4. Liu J, Li Y, Zhao D, et al. Efficacy and safety of brain-computer interface for stroke rehabilitation: an overview of systematic review. Front Hum Neurosci. 2025; 19:1525293.
5. Leuthardt EC, Wilk SJ, Souders L, et al. At-home BCI rehabilitation therapy for chronic upper extremity deficit after stroke (BCI-REHAB): a randomized trial. Stroke Vasc Interv Neurol. 2026; 6.
6. Mortezaei A, Al-Saidi N, Taghlabi KM, et al. Brain-computer interfaces in poststroke rehabilitation: a meta-analysis of randomized clinical trials. Neurosurg Focus. 2026; 60(2).
7. Prasad NK, Perry NJ, Goldring AL, et al. A retrospective analysis of post-stroke rehabilitation with real world use of brain-computer interface. J Neuroeng Rehabil. 2026; 23(1):68.
8. Rustamov N, Humphries J, Carter A, et al. Theta-gamma coupling as a cortical biomarker of brain-computer interface-mediated motor recovery in chronic stroke. Brain Commun. 2022; 4(3):fcac136.
9. Rustamov N, Souders L, Sheehan L, et al. IpsiHand brain-computer interface therapy induces broad upper extremity motor rehabilitation in chronic stroke. Neurorehabil Neural Repair. 2025; 39(1):74-86.

Government Agency, Medical Society, and Other Authoritative Publications:

1. Food and Drug Administration (FDA). Neurolutions IpsiHand Upper Extremity Rehabilitation System (DEN200046). Available at: https://www.accessdata.fda.gov/cdrh_docs/pdf20/DEN200046.pdf. Accessed on August 7, 2026.
2. Ifejika NL, Awosika OO, Black T, et al. Improving access to stroke rehabilitation and recovery: a policy statement from the American Heart Association/American Stroke Association. Stroke. 2025; 56(9).
3. Management of Stroke Rehabilitation Work Group. VA/DoD clinical practice guideline for management of stroke rehabilitation. Version 5.0. US Department of Veterans Affairs; US Department of Defense. May 2024. Available at: https://www.healthquality.va.gov/guidelines/Rehab/stroke/index.asp. Accessed on August 20, 2026.

1. American Stroke Association. Stroke Rehabilitation. Available at: https://www.stroke.org/en/life-after-stroke/stroke-rehab. Accessed on August 7, 2026.

Electroencephalography-driven upper extremity powered exerciser
IpsiHand

The use of specific product names is illustrative only. It is not intended to be a recommendation of one product over another, and is not intended to represent a complete listing of all products available.

Federal and State law, as well as contract language, including definitions and specific contract provisions/exclusions, take precedence over Medical Policy and must be considered first in determining eligibility for coverage. The member’s contract benefits in effect on the date that services are rendered must be used. Medical Policy, which addresses medical efficacy, should be considered before utilizing medical opinion in adjudication. Medical technology is constantly evolving, and we reserve the right to review and update Medical Policy periodically.

No part of this publication may be reproduced, stored in a retrieval system or transmitted, in any form or by any means, electronic, mechanical, photocopying, or otherwise, without permission from the health plan.

© CPT Only – American Medical Association