Linus Health Presentations at AAIC 2025

Key takeaways:

People value different outcomes when it comes to brain health, highlighting the need for care that reflects personal goals, not just clinical symptoms.
Using a language model architecture, we effectively clustered 16,000 free-text responses into 545 robust and meaningful themes, capturing the diversity in individuals’ brain health priorities.
This approach opens the door to personalized brain health interventions and education at scale by linking individuals’ unique goals to actionable support strategies.

Key takeaways:

The 7-minute digital assessment of cognition (DAC) protocol achieved excellent accuracy in distinguishing individuals with MCI from cognitively unimpaired patients.
By combining traditional outcome measures with process-based and acoustic features, DAC identified subtle differences often missed by traditional tests.
The study supports rDAC as a sensitive, remote-friendly alternative to time-consuming neuropsychological evaluations, helping enable earlier detection and timely intervention for individuals at risk of dementia.

Key takeaways:

The 3-minute Digital Clock and Recall (DCR) assessment accurately identified cognitive impairment and predicted amyloid PET status, matching or outperforming traditional tests and blood-based biomarkers.
When combined with blood-based biomarkers, the DCR significantly boosted amyloid-beta prediction accuracy to levels comparable with CSF biomarkers.
This scalable, AI-enabled tool offers a cost-effective way to identify eligible participants for AD clinical trials, helping reduce screening failures and accelerate access to disease-modifying treatments.

Key takeaways:

The 7-minute Digital Assessment of Cognition (DAC) protocol showed high concordance with traditional paper-and-pencil neuropsychological assessments across multiple cognitive domains.
DAC metrics, particularly memory and executive function indices, were strongly associated with traditional measures used to classify subtypes of MCI and dementia.
These findings support the utility of DAC as a valid, scalable, and easy-to-administer tool for detecting early cognitive impairment in clinical and remote settings.

Key takeaways:

A multimodal model combining Digital Clock and Recall (DCR), digital Trail-Making Test - Part B (dTMT-B), blood-based biomarkers, and APOE status accurately predicted Aβ-PET positivity, with performance approaching that of CSF-based methods.
The model, capturing drawing, speech, and task performance metrics, offers an efficient way to identify patients likely to be amyloid-positive on a PET scan.
This approach supports earlier identification of AD patients who are likely to be suitable candidates for AD clinical trials and disease-modifying treatments.

Key takeaways:

The Linus Health digital Trail Making Test-Part B (dTMT-B) successfully distinguished MCI from normal cognition using detailed process metrics such as total/hit duration, pen stroke velocity/efficiency, and drawing efficiency.
Novel indices derived from dTMT-B helped separate neurocognitive delays from motor impairments, providing a clearer picture of what drives test performance.
These findings highlight the potential of dTMT-B to more accurately distinguish between motor and cognitive causes of slowed performance, improving early detection of cognitive impairment and reducing misclassification.

Key takeaways:

The Digital Clock and Recall (DCR) showed strong concordance with widely used assessments, including the NIH Toolbox, TICS, ACE-R, and MMSE, in a large UK-based study.
The DCR stood out among all the other cognitive assessments in its association with AD diagnosis.
This is the first evidence from a large, representative UK cohort confirming DCR’s clinical utility, supporting its broader adoption as a brief, automated tool for detecting cognitive impairment and Alzheimer’s disease.

Key takeaways:

An abbreviated version of the Linus Health digital Trail Making Test-Part B (dTMT-B) successfully distinguished individuals with MCI from those with normal cognition, using only the first half of the test.
Longer time spent inside target circles was associated with lower scores in attention, working memory, and episodic memory, highlighting its sensitivity to early cognitive decline.
The abbreviated dTMT-B is easy to administer and shows promise as a rapid, process-based digital tool for frontline identification of emergent cognitive impairment.

Learn about the latest study findings from Linus Health