Brain-Computer Interface Communication Device for People With ALS

Within-subject longitudinal evaluation. Each participant serves as their own control and uses all three modalities across a 12-week primary endpoint window: (a) their existing commercial eye-gaze speech generating device as the active comparator, (b) the Cognixion ONE AR-BCI configuration, and (c) the Cognixion ONE screen-based BCI configuration. Letterboard is retained as a within-subject fallback for participants whose eye-gaze becomes non-viable during the window.

The Cognixion ONE platform combines a constant medical-grade EEG sensing assembly (Axon-R, IEC 60601 certified) with one of three interchangeable display configurations - an Apple Vision Pro augmented-reality headset, XReal augmented-reality glasses, or an external screen - and a Cognixion-developed AI-assisted communication application. Selection is made by steady-state visual evoked potential response, optionally combined with eye tracking as a hybrid modality. Display configuration is selected per participant and may be changed during the study under documented criteria with participant re-consent.

All study activity is conducted by Cognixion personnel in participants' homes, with weekly remote supervision and at-least-monthly on-site visits. Recruitment partners provide referrals only and do not execute study activities. Sessions run 3 to 5 times per week at approximately one hour each, with a trained caregiver present for all sessions using the study device.

Enrollment is rolling across the first 18 months of a 36-month period of performance. Disease progression is tracked by the complete 12-item ALSFRS-R at baseline and every 4 weeks, entering the primary statistical model as a time-varying covariate.

The study is funded by the National Institute on Deafness and Other Communication Disorders under SBIR award R44DC023888, and builds on a completed Phase 1 feasibility study (NCT07209943).

• Cognixion ONE AR-BCI configuration (Device): Non-invasive EEG-based brain-computer interface using steady-state visual evoked potentials, presented on a head-worn augmented-reality display, with AI-assisted word and phrase suggestion. Input modalities available: pure BCI, hybrid eye-tracking plus BCI, eye tracking, and switch control, subject to display configuration.
• Cognixion ONE screen-based BCI configuration (Device): The same EEG sensing assembly and communication application presented on an external screen rather than a head-worn display.
• Existing eye-gaze speech generating device (Device): The participant's own prescribed commercial eye-gaze speech generating device, used as the active within-subject comparator.

Primary outcomes

• Voluntary daily use time per input modality [12 weeks]: Minutes per day of voluntary use of each of the three communication modalities, captured from device telemetry. Measures user preference across available options, and is robust to learning effects, communication-partner effects, and prompt-driven artifacts that bias single-session performance measures. Minmum:0mins/day Maximum: 1440mins/day

Secondary outcomes

• Communication rate [12 weeks]: Words per minute during structured communication tasks and free conversation, per modality.
• Selection accuracy [12 weeks]: Percent correct selections on the BCI Game ground-truth task. Benchmark: at least 80 percent accuracy on at least two BCI targets. Range: 0-100%
• Information Transfer Rate [12 weeks]: Bits per minute, computed from BCI Game performance and from communication-task throughput. Minimum: 0 bits/min, no maximum
• Modality preference and time allocation [12 weeks]: Fraction of voluntary use time spent in each input mode: pure BCI, hybrid eye-tracking plus BCI, eye tracking, and switch control. Normalized value: 0-1.0 for each modality.
• System Usability Scale (SUS) [Baseline and multiple longitudinal time points through 12 weeks]: Standardized 10-item usability instrument. Score range 0 to 100; higher scores indicate better usability.
• NASA Task Load Index (NASA-TLX) [12 weeks]: Six-subscale cognitive workload assessment administered after structured task sessions. Higher scores indicate greater workload. Score range: 0-100
• ALS-Specific Quality of Life Instrument, Short Form (ALSSQOL-SF) [Baseline and 12 weeks]: Validated ALS-specific quality of life measure. Higher scores indicate better quality of life. Range: 0-10
• Zarit Burden Interview (ZBI) [Baseline and 12 weeks]: Caregiver burden assessment administered to the designated caregiver. Higher scores indicate greater burden. Score range: 0-88
• Conversation metrics [12 weeks]: Turn-taking dynamics, dialog efficiency, AI suggestion-acceptance rate, and vocabulary diversity (type-token ratio). Normalized scores: 0-1
• Disease progression (ALSFRS-R) [Baseline and every 4 weeks through 12 weeks]: Complete 12-item ALS Functional Rating Scale - Revised. Score range 0 to 48; higher scores indicate better function. Evaluator-administered at baseline, validated self-entry version thereafter. Score range: 0-48

Other outcomes

• Sustained pure-BCI use after eye-gaze decline [12 weeks]: Continued use of pure-BCI input among participants whose eye-gaze function declines during the study. Observational
• Caregiver setup time [12 weeks]: Time for the caregiver to set up the device for a session. Target is 5 minutes or less by end of cohort onboarding. Minimum is 1min, no maximum.
• Author-of-Record disclosure compliance [12 weeks]: Rate of participant confirmation on AI-suggested phrases prior to the phrase being spoken. Normalized score: 0-1
• Articulatory precision trajectory [12 weeks]: Within-person rate of change in articulatory precision from a weekly recorded reading passage, and its concordance with the ALSFRS-R speech item, bulbar subdomain, and total score. Analyzed within the speech-analytic evaluable subset. Normalized: 0-1
• Time to loss of measurable speech [12 weeks]: Time from enrollment to the last scorable speech session, within the speech-analytic evaluable subset. Censored at Week 12 or withdrawal. Days from start of study.

Sex: All, ages: 18 Years to any

Inclusion Criteria:

* Diagnosis of ALS established by a neurologist, meeting either the Gold Coast Criteria (preferred) or the revised El Escorial criteria (definite, probable, probable laboratory-supported, or possible ALS), documented by the referring or treating neurologist
* ALSFRS-R Speech subscale score of 3 or lower (impaired but functional speech) through 0 (anarthric)
* Documented use of a Tobii Dynavox or Smartbox eye-gaze speech generating device as primary or secondary communication method for at least 30 days prior to screening. Candidates who meet all other criteria but lack a prescribed eye-gaze device may be provided a loaner device, capped at 25 percent of the enrolled cohort
* Age 18 years or older
* A designated, trainable caregiver able to be present for structured study sessions and to complete a 2-hour onboarding and a Month-3 refresher
* Volitional eye opening sustained independently for at least 30 minutes, and sufficient visual function to perceive the augmented-reality display stimuli
* A reliable communication method other than the study device for affirmative assent, at minimum a yes/no channel via residual vocalization, head movement, blink, or brow raise
* Sufficient English comprehension to engage with the study protocol and the AI conversational interface

Exclusion Criteria:

* Active hospice enrollment without documented capacity to participate in research activities
* Cognitive or behavioral impairment of a degree that would preclude successful participation in study procedures, as judged by the treating neurologist, a study neurologist, or the study speech-language pathologist
* History of photosensitive epilepsy or active seizure disorder
* Sensitivity to flashing lights
* Severe uncorrected visual impairment that precludes use of an augmented-reality display
* Vision disorders restricting the visual field, such as glaucoma with significant field loss, diplopia, or nystagmus
* History of vertigo or other vestibular disorders that would be aggravated by augmented-reality display use
* Claustrophobia related to head-mounted display use, after attempted accommodation
* Scalp condition prone to irritation, inflammation, injury, or infectious process that precludes safe placement of EEG electrodes
* Severe oculomotor impairment that precludes establishment of an initial eye-gaze calibration on the participant's existing eye-gaze device
* Concurrent enrollment in another investigational communication-device or brain-computer interface trial
* Severe hearing impairment or deafness that would prevent participation in conversational tasks

• Wolpaw JR, Bedlack RS, Reda DJ, Ringer RJ, Banks PG, Vaughan TM, Heckman SM, McCane LM, Carmack CS, Winden S, McFarland DJ, Sellers EW, Shi H, Paine T, Higgins DS, Lo AC, Patwa HS, Hill KJ, Huang GD, Ruff RL. Independent home use of a brain-computer interface by people with amyotrophic lateral sclerosis. Neurology. 2018 Jul 17;91(3):e258-e267. doi: 10.1212/WNL.0000000000005812. Epub 2018 Jun 27. PMID 29950436
• Cedarbaum JM, Stambler N, Malta E, Fuller C, Hilt D, Thurmond B, Nakanishi A. The ALSFRS-R: a revised ALS functional rating scale that incorporates assessments of respiratory function. BDNF ALS Study Group (Phase III). J Neurol Sci. 1999 Oct 31;169(1-2):13-21. doi: 10.1016/s0022-510x(99)00210-5. PMID 10540002

• Chris Ullrich, M.Sc., Principal investigator, Cognixion Corporation

NIH grants: R44DC023888