At the 2025 Alzheimer’s Association International Conference (AAIC) in Toronto, digital health company Linus Health is making its case for a new era in cognitive care—one powered by artificial intelligence (AI), rapid digital assessments, and a more personalized approach to brain health.

The Boston-based startup is presenting a wide slate of research at this year’s AAIC, including two oral presentations and eight scientific posters. Across these studies, one message comes through clearly: traditional approaches to cognitive testing may soon give way to digital tools capable of detecting Alzheimer’s disease and related dementias earlier, faster, and more cost-effectively.

“The studies we’re presenting represent the next wave of translational science in brain health,” said Dr David Bates, co-founder and CEO of Linus Health. “By combining validated digital tools with AI, we’re enabling earlier and more accurate detection of Alzheimer’s and related conditions—while also expanding access to care through at-home testing options.”

Rethinking Recruitment for Alzheimer’s Trials

Among the most closely watched data at the conference is a study led by Linus Health’s director of cognitive science, Dr. Ali Jannati, focused on improving recruitment for Alzheimer’s clinical trials. Using a brief digital cognitive test called Digital Clock and Recall (DCR), combined with a second task, the digital Trail Making Test Part B (dTMT-B), researchers trained machine-learning models to identify cognitive impairment and amyloid-beta PET status—key criteria for Alzheimer’s trials.

The results, presented in both a poster and a dedicated “Developing Topics” session, show the model’s predictions aligned closely with more invasive and costly cerebrospinal fluid tests. When combined with blood biomarkers and APOE genotype data, the tool approached the accuracy of gold-standard diagnostics.

Such a model could help streamline pre-screening, reducing screening failures and accelerating access to disease-modifying trials.

Personalized Brain Health—at Scale

Beyond diagnostics, Linus is also making a push for more individualized care. One poster presentation described how researchers used a language model to analyze over 16,000 responses about personal goals related to brain health. The algorithm sorted the data into 545 distinct themes—ranging from maintaining independence to reducing memory lapses—illustrating just how varied people’s cognitive health priorities can be.

“That kind of granularity allows us to go beyond a one-size-fits-all approach,” Jannati said. “We can begin to develop personalized interventions based on what really matters to each person.”

Another study introduced a seven-minute digital cognitive assessment (DAC) capable of classifying individuals into distinct cognitive subgroups, including amnestic and dysexecutive types of mild cognitive impairment (MCI), using semi-supervised machine learning. According to the company, this work could enable more tailored patient stratification in both research and care settings.

Many of Linus Health’s tools are designed for remote use, including the DAC protocol, which achieved high accuracy in distinguishing MCI from normal cognition using traditional outcome scores as well as speech, drawing, and timing data.

In separate research presented by Dr. David Libon, the company’s scientific advisor and a neuropsychologist, digital metrics from the dTMT-B were found to be sensitive not only to cognitive decline but also capable of distinguishing between neurocognitive and motor impairments.

That kind of resolution, Libon argues, has long been lacking in quick, scalable cognitive tests.

“For decades, we’ve known that how someone performs a test can be as important as whether they get the answer right,” Libon said. “These digital tools finally let us quantify that process—and scale it beyond the neuropsychologist’s office.”

Libon’s team also unveiled an abbreviated version of the dTMT-B that preserved its diagnostic power using only the first half of the test, a potential time-saver for primary care settings.

Evidence from the Field

Linus Health supports its platform with real-world data. In a UK clinical study, researchers evaluated DCR’s performance against popular assessments such as the Mini-Mental State Examination (MMSE), Addenbrooke’s Cognitive Examination – Revised (ACE-R), and the NIH Toolbox, all of which are cognitive tests with varying focuses. The digital assessment closely matched these traditional tools and proved to be the most accurate predictor for Alzheimer’s diagnosis within the group.

As per the company, the finding supports broader international adoption of DCR as a front-line screening method for cognitive impairment. Linus Health is betting that AI-enhanced assessments will become essential tools in both clinical research and routine care. Its digital platform is already in use by healthcare systems, life sciences companies, and research institutions.

At this year’s AAIC, the company also previewed forthcoming enhancements to its research platform, including new features that allow patients and study participants to complete assessments from home.

That remote capability, Bates said, could be crucial for reaching populations historically underserved in brain health research and care.

“Too often, we wait until cognitive decline becomes a crisis,” Bates said. “Our vision is to make brain health something people manage proactively throughout life—and digital tools are the key to making that vision a reality.”

As Alzheimer’s management landscape continues its shift toward earlier diagnosis and preventive strategies, Linus Health is positioning itself at the nexus of neuroscience, AI, and digital health, with the intention of helping more people live longer, healthier lives with better cognitive outcomes.