8 hours ago
Linus Health Digital Cognitive Assessments Meet or Surpass New CEOi Benchmarks for Early Cognitive Impairment Detection
Published and emerging evidence shows strong alignment across detection, diagnostic support, and cognitive profiling—while supporting faster, more actionable brain health workflows
BOSTON, May 27, 2026 -- Linus Health, an AI-driven brain health company focused on early detection and proactive intervention, announced that its digital solutions meet or exceed the performance criteria suggested in recent recommendations from the Global CEO Initiative on Alzheimer's Disease (CEOi) for clinic-based digital cognitive assessments.
Published in Alzheimer's & Dementia, the CEOi framework defines three core clinical use cases for supervised, in-clinic digital cognitive assessments (DCAs):
- Initial detection of cognitive impairment (Detection DCA);
- Diagnostic support for mild cognitive impairment (MCI) and dementia (Diagnostic DCA); and
- Cognitive profile characterization to aid in subtyping or etiologic workup (Profile Characterization Aid DCA).
The workgroup calls for digital tools to at least match, and ideally improve upon, a set of ambitious performance measures, while fitting real-world clinical workflows.
Direct comparison to the CEOi detection benchmark
Within the CEOi Detection DCA benchmarks—approximately 3 to 5 minutes, 80% sensitivity, and 85% specificity—Linus Health's remote-enabled Digital Assessment of Cognition (DAC) exceeded the suggested benchmark in an independent test set, achieving 90.7% sensitivity and 100% specificity for distinguishing cognitively unimpaired individuals from those with cognitive impairment, including MCI and dementia.
The DCR, a <3-minute assessment, achieves strong performance in brief cognitive evaluations, with 80% sensitivity and 73% specificity for general cognitive impairment and 81% sensitivity and 80% specificity for verbal memory impairment. DCR studies showed higher sensitivity and overall accuracy compared with MMSE, Mini-Cog, and MoCA (based on research planned for presentation at the Alzheimer's Association International Conference [AAIC] 2026), directly supporting the CEOi recommendation that brief DCAs match or outperform legacy non-digital screening tools.
Strong alignment in diagnostic-aid workflows
Within the CEOi Diagnostic Aid DCA benchmarks—approximately 10 to 20 minutes, 85% sensitivity and 90% specificity for MCI detection—Linus Health's DAC achieved 82.9% sensitivity and 100% specificity for distinguishing cognitively unimpaired individuals from those with MCI, with a positive predictive value of 100% and negative predictive value of 78% in the study sample. The same DAC research showed 73.7% sensitivity and 85.7% specificity for distinguishing MCI from dementia. These data show that Linus Health meets the CEOi specificity benchmark for diagnostic support and approaches the suggested sensitivity threshold, while delivering DAC in about seven minutes—shorter than the CEOi's suggested 10- to 20-minute diagnostic aid window.
Evidence also supports profile characterization and etiologic workup
For the CEOi Profile Characterization Aid DCA use case, the guidelines describe a more comprehensive tool intended to support etiologic workup and uses 85% sensitivity and 90% specificity as an example benchmark for detecting MCI due to Alzheimer's disease. Linus Health is one of the only DCAs tackling this critical third use case. Linus' evidence base supports that goal in two ways:
- DAC demonstrates the ability to differentiate clinically meaningful phenotypes, including amnestic, dysexecutive, and multi-domain MCI subtypes; has greater than 90% concordance with classifications derived from 2- to 3-hour comprehensive neuropsychological evaluations; and achieved (page 290) a 93% negative predictive value for ruling out p-tau217 positivity in real-world settings.
- DCR predicts amyloid PET positivity with an AUC of 0.81. In a separate DCR analysis focused on identifying patients with MCI or mild dementia likely due to Alzheimer's disease, a combination of DCR's cognitive-impairment and amyloid-prediction models achieved (page 289) 83% sensitivity and 85% specificity for identifying that cohort.
These findings support the CEOi framework's emphasis on understanding not only whether cognitive impairment is present, but also what pattern of impairment is emerging and whether Alzheimer's pathology may be more likely.
"Clear, clinically relevant standards are essential for moving digital cognitive assessment from promise to routine practice," said David Bates, PhD, CEO and co-founder of Linus Health. "The CEOi's recommendations reinforce the direction we have believed in from the beginning: brief, scalable, scientifically grounded tools that deliver more actionable information than legacy screening approaches and help patients get to the right next step sooner. Merely identifying patients with dementia is insufficient for the future of brain health. Patients deserve more."
Together, these results support the use of Linus Health solutions as rapid, low-burden tools that provide primary care physicians with what they need to determine which patients require deeper workup, specialist referral, or longitudinal monitoring.
"Digital cognitive assessments should do far more than recreate paper tests on a screen," said Alvaro Pascual-Leone, MD, PhD, co-founder and chief medical officer of Linus Health. "They should capture richer behavioral signals, help clinicians interpret what those signals mean, and support better decisions about who needs follow-up, which patients may be at highest risk, and how care can be tailored earlier in the disease course."
By separating standards for detection, diagnostic support, and profile characterization, the CEOi recommendations provide health systems, clinicians, researchers, and developers with a more practical framework for evaluating clinical utility. Linus Health believes the recommendations will help accelerate broader adoption of more sensitive, scalable, and clinically meaningful brain health tools—and underscore the company's continued focus on evidence-based solutions that improve earlier identification and intervention.
About Linus Health
Linus Health is a Boston-based digital health company focused on improving brain health around the world. The company develops science-driven digital assessments and AI-enabled analytics that help clinicians and researchers identify cognitive change earlier, guide next steps in care, and support more proactive, personalized intervention. Linus Health works with healthcare delivery organizations, research institutions, and life sciences partners to advance earlier detection and better brain health outcomes.
Media Contact
Supreme Communications for Linus Health
linushealthPR@supremecomms.ai