For decades, the standard for checking a patient’s cognitive health has been the Mini-Mental State Examination (MMSE) or the Mini-Cog—brief, pen-and-paper tests that are notoriously blunt instruments. They are often too slow to administer in a busy primary care office and too imprecise to catch early-stage impairment.

That is changing. A new framework from the Global CEO Initiative on Alzheimer's Disease (CEOi), published in Alzheimer's & Dementia, has set ambitious performance targets for digital cognitive assessments (DCAs). Now, data from Linus Health suggests that AI-driven digital tools are not just meeting these new standards—they are outperforming the legacy tests that have dominated clinical practice since the 1970s.

The Shift to Digital Precision

The CEOi framework establishes three distinct tiers for digital brain health tools: initial detection, diagnostic support, and profile characterization. The goal is to move beyond simple 'yes or no' screening toward actionable insights that help clinicians determine whether a patient’s memory lapses are normal aging or the early signs of Alzheimer’s pathology.

Linus Health’s Digital Assessment of Cognition (DAC) was tested against these new benchmarks. In an independent test set, the tool achieved 90.7% sensitivity and 100% specificity for distinguishing cognitively unimpaired individuals from those with impairment. For context, the CEOi benchmark for detection is 80% sensitivity and 85% specificity. By clearing these hurdles, the platform demonstrates that digital tools can provide a higher degree of clinical confidence in a fraction of the time.

Beyond Simple Screening

The most significant hurdle in brain health is not just identifying impairment, but understanding its cause. While legacy tests like the MoCA (Montreal Cognitive Assessment) provide a snapshot of current function, they rarely offer insight into the underlying biology.

Linus Health is attempting to bridge this gap by tackling the CEOi’s third use case: profile characterization. Their data shows a 93% negative predictive value for ruling out p-tau217 positivity—a key biomarker for Alzheimer’s—in real-world settings. Furthermore, their Digital Cognitive Review (DCR) can predict amyloid PET positivity with an AUC of 0.81.

"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. "Merely identifying patients with dementia is insufficient for the future of brain health. Patients deserve more."

Why the Timing Matters

The clinical workflow is currently the biggest bottleneck in Alzheimer’s care. Primary care physicians often lack the time for comprehensive neuropsychological evaluations, which can take up to three hours. By providing a digital alternative that takes roughly seven minutes to administer, Linus Health is positioning its tools to fit into the standard 15-minute primary care visit.

This efficiency is critical as new disease-modifying therapies for Alzheimer’s reach the market. These drugs are most effective when administered early, making the speed and accuracy of initial detection the primary factor in patient outcomes.

Key Takeaways

  • Performance Gains: Linus Health’s DAC exceeded CEOi benchmarks for detection, achieving 90.7% sensitivity and 100% specificity in independent testing.
  • Workflow Efficiency: The platform provides diagnostic-level insights in approximately seven minutes, significantly faster than the 10- to 20-minute window suggested by the CEOi.
  • Biological Insight: Unlike legacy paper tests, these digital tools show strong concordance with comprehensive neuropsychological evaluations and can help predict the presence of Alzheimer’s-related biomarkers.

What Experts Say

Clinical adoption of these tools will likely hinge on the upcoming presentation of data at the Alzheimer's Association International Conference (AAIC) 2026. If the research confirms that digital tools consistently outperform legacy tests like the MMSE and Mini-Cog in diverse clinical populations, the pressure on health systems to transition to digital-first brain health workflows will intensify.

For clinicians, the next decision point arrives this summer. As the AAIC 2026 data becomes public, the focus will shift from whether these tools can work to whether they can be integrated into the electronic health records of large hospital systems without adding administrative burden. If the integration proves seamless, the era of the paper-based cognitive screen may finally be coming to a close.

This article is for informational purposes only. Always consult a qualified healthcare professional before making any medical decisions.