Tolion Brain Coach: AI-Enabled Personalization for Brain Health Prevention

Lerch, O.¹, Veverova, K.¹, Kadera, P. ¹, Katonová, A., Chmátalová, Z., Zahálka, J. ¹, Prucek P, Kodejs Z. Tolar, M.¹

¹ Tolion Health, The Blox, Evropská 11/2758, 160 00 Praha 6, Czech Republic

Background

The preclinical phase of Alzheimer's disease (AD) precedes the onset of clinical symptoms by decades, creating a substantial window for preventive interventions aimed at modifying disease trajectory. Recent evidence suggests that up to 45% of dementia cases may be attributable to modifiable risk factors, with structured and sustained lifestyle interventions demonstrating greater effectiveness than non-structured approaches. Despite this potential, scalable tools enabling individualized, evidence-based prevention strategies remain limited. Tolion Health AI developed the Tolion Brain Coach, a digital platform that delivers personalized brain health guidance grounded in current scientific evidence to support individuals in reducing their long-term risk of dementia.

Methods

The Tolion Brain Coach (i) integrates multimodal individual-level data, including continuous device-derived health metrics obtained through Apple HealthKit and Health Connect (covering sleep, physical activity, cardiovascular fitness, autonomic function, and recovery), self-reported information from structured questionnaires, AI-supported conversational input, and, where available, genetic risk status. The platform (ii) computes an individualized, weighted Brain Coach Score that combines fifteen modifiable and non-modifiable metrics into a single, interpretable measure of brain health. Using this profile, an AI conversational coach (iii) generates tailored, action-oriented recommendations targeting modifiable risk factors and reinforces science-backed daily habits. For carriers of high-risk genotypes, a dedicated APOE4 Protection Mode reframes guidance around protective behaviors. The underlying models account for variation across sex, age, comorbidities, and medication use, enabling individualized rather than population-level recommendations.

Results

The platform enables personalized risk profiling and continuous tracking of the Brain Coach Score, illustrating how targeted modifications in lifestyle and health behaviors are reflected in a user's brain health trajectory over time. Case-based use demonstrates meaningful heterogeneity in personalized recommendations across individuals, underscoring the limitations of one-size-fits-all prevention approaches.

Conclusion

AI-supported personalization may represent a scalable approach to dementia risk reduction by translating complex, multidimensional health data into individualized, actionable preventive guidance that users can engage with daily.