Neurotrack

Deliberate C, placing this alongside BrainCheck at the low-AI end of the cognitive assessment spectrum. The current clinician products are described as patient-led tests selected for validity, accessibility and ease of administration, with instant objective scoring against DSM-5 aligned criteria. That is standardised digital delivery and automated scoring of validated instruments, not a machine learning model deriving novel signal.

Notably the company's own historical eye-tracking work, which involved genuine digital phenotyping and would have graded higher, has been REMOVED from the clinician product suite by the company's own statement. The moat here is workflow fit, deployment speed and health system distribution rather than the algorithm.

Screening tool producing objective scores that route patients toward clinician evaluation and care pathways rather than diagnosing. Positioned explicitly to help care teams work at the top of their license and to reduce provider burden, with the test administrable by staff without specialised training.

Graded B rather than A because no explicit stand-alone-diagnostic disclaimer of the kind BrainCheck and Cognivue both publish prominently was located, and DSM-5 aligned diagnostic scoring language sits closer to the diagnostic line than a pure screening framing would.

Credit specifically for a correction the company makes about ITSELF, which is rare and worth noting. Neurotrack originated in eye-tracking cognitive assessment and much third party coverage, including recent write-ups, still describes it that way. The company states plainly on its own site that while digital eye tracking was one of its first technologies, it is not part of its current clinician product suite.

Publicly correcting an outdated and more impressive-sounding description of your own technology is the opposite of the usual incentive. Graded B rather than A because the specific instruments now comprising the three minute battery are not enumerated, so a buyer cannot tell which validated tests are being delivered.

Nothing identifies any party in the chain: no model or model family, no hosting provider, no sub processor list was located in two passes, and no retention policy or statement on secondary use of assessment data was found. Compliance attestations are stated directly alongside the product rather than buried, which is worth noting as a presentation choice even though it names no one. The content is worth describing precisely because it is easy to underestimate.

Cognitive assessment results are among the most consequential health data a person can generate outside a diagnosis: a recorded impairment signal reaches into driving, employment, financial capacity and legal capacity, and unlike most test results it is frequently generated for people who are already worried about exactly those consequences.

Where that data sits, who else can reach it, and whether it contributes to a normative or training corpus are therefore questions with unusual weight, and none is answered. This index recorded the same normative corpus question for a peer in this category, where the reference population quietly shifts as clinical results accumulate. Ask for the hosting arrangement, a sub processor list, the retention schedule, and whether assessment results contribute to any reference dataset or model development.

Substantial and long-running. 25 peer reviewed publications, 11 patents, and grant funding from the National Institute on Aging, the Alzheimer's Drug Discovery Foundation including a $750,000 award to evaluate longitudinal cognitive monitoring, and the Georgia Research Alliance, which is non-dilutive scientific validation rather than marketing spend.

Founders and clinical team collectively report 354 peer reviewed papers cited more than 37,000 times, with genuine standing in memory and cognitive neuroscience. The platform is stated to meet or exceed Alzheimer's Association criteria for brief cognitive assessments.

Graded B rather than A for a specific reason: a substantial share of that publication record derives from the eye-tracking research lineage that the company has since removed from its clinician products, so the evidence base and the shipping product are not fully aligned. Sensitivity and specificity claims for the current three minute test were not located in retrievable published form.

SOC 2 and HIPAA compliance stated directly alongside the product rather than buried, and the cloud platform handles assessment data across health system deployments. Graded B rather than A because no published retention policy or statement on secondary use of assessment data was located.

HIPAA compliance stated explicitly and prominently as a product attribute. Graded B rather than A because no BAA terms or execution detail were located, but this is a clearer published position than most vendors in this category offer.

Neurotrack publishes SOC 2 compliance directly on its product page. Graded B rather than A because the disclosure does not specify Type I versus Type II, which is the distinction that carries the weight: a Type I report covers the design of controls at a single point in time, while a Type II covers their operating effectiveness over a period. No trust centre or downloadable report was located, and no other attestation was found.

The omission is more consequential now that the rest of this category has been assessed against the same axis. A direct peer states SOC 2 Type 2 explicitly. Another holds HITRUST certification together with a TX-RAMP Level 2 government cloud authorisation. A third publishes a device security document naming individual controls including its firewall posture, antivirus and role separation. Neurotrack's disclosure is no longer the most specific in the group, and the single word it omits is the word that would settle where it belongs.

Buyers should ask which SOC 2 type the report covers, request the report itself under a non disclosure agreement, and ask when the audit period closed and when the next one is scheduled.

Consistently described as FDA registered as a Class II medical device, which is a materially weaker statement than clearance and the difference matters. Registration and listing mean a manufacturer has notified the FDA of its establishment and devices; clearance means the FDA reviewed a submission and found substantial equivalence to a predicate. No 510(k) number or De Novo was located for Neurotrack.

This is precisely the terminology ambiguity that Cognivue's documented De Novo DEN130033 resolves for this category, and buyers comparing the two should understand they are not equivalent regulatory positions despite similar-sounding language.

Graded C rather than lower because Class II registration is a real regulatory status, the company uses the accurate word registered rather than claiming clearance, and the platform is stated to meet or exceed Alzheimer's Association criteria for brief cognitive assessments.

Addressed by design rather than by study, which distinguishes it from Cognivue's empirical FOCUS approach. The tests deliberately use culturally agnostic symbols and numbers, stated explicitly as a choice to ensure inclusivity and remove barriers to care, and the product requires no specialised training to administer, which reduces administrator-driven variation. Designing out language and cultural dependency is a legitimate mitigation for the central bias risk in cognitive testing. Graded B rather than A because no published subgroup performance data was located, so the design intent is not backed by measured results the way Cognivue's stratified FOCUS study is.

One act of candour on this record is rare enough to name specifically, because it runs against the usual incentive. The company originated in eye tracking cognitive assessment, and much third party coverage still describes it that way, which is the more impressive sounding technology. Its own site states plainly that while digital eye tracking was one of its first technologies, it is not part of the current clinician product suite.

Publicly correcting an outdated and more flattering description of your own capability is the opposite of what this index usually finds, where vendors let generous third party characterisations stand uncorrected and sometimes cite them. It also has practical value: a buyer who read the coverage and expected eye tracking is told before they buy rather than after. Held at C because what replaced it is not enumerated.

The instruments now comprising the short battery are not listed, so a buyer cannot tell which validated tests are being delivered, and in cognitive assessment the identity of the instrument is the evidence, since each one has its own literature, its own normative data and its own known limitations. No accuracy figure for the impairment output, no evaluation methodology and no warranty, indemnity or remediation commitment was located. Ask which instruments the battery contains, what normative data the scoring uses, and what a clinician sees about the confidence of a screening result.

The strongest interoperability position in the cognitive assessment category, and the reason this vendor completes the comparison set. The platform is stated to integrate with all major EHR systems, with EHR integration achieved from 2023 onward across health delivery organisations, and deployment measured in weeks rather than months.

Assessment results flow into clinical workflow and support Medicare Annual Wellness Visit compliance and RADV audit documentation, meaning the output lands in the systems where quality and risk adjustment work actually happens. Against Cognivue, graded C here as a standalone hardware device generating a paper report, this is the opposite end of the category.

Cloud based, runs on any device, requires no specialised hardware or training, and supports in-clinic, virtual and in-home administration with remote and population-wide screening capability. Deploys in weeks. Used by health systems, health plans, home health agencies and primary care alike. This is the most flexible deployment model in the category, against Cognivue's dedicated hardware and Linus Health's iPad plus Apple Pencil requirement. No data residency disclosure located, which is the only qualification.

No pricing published and no pricing basis disclosed. The company does signal the buyer's economic case clearly, aligning with value-based care models, Medicare Annual Wellness Visit compliance, RADV audit documentation and Medicare Advantage Star Ratings strategy, so the return argument is legible even though the cost is not. Graded C rather than higher because, unlike BrainCheck, no specific CPT codes or reimbursement guidance are published.

The broadest deployment footprint in this category. Spans primary care, specialty settings, virtual care and care at home, serving health systems, health plans, home health agencies and provider groups, with named customers including Sutter Health, Providence and Stanford Medicine and reported reach across more than 15,000 clinicians.

The product line extends past screening into patient education addressing modifiable risk factors, care management tooling and a cognitive behavioural health digital therapeutic. Serving the screening, monitoring and risk-factor intervention stages across both clinical and payer channels is materially wider than a single-assessment vendor.