AI Integration in Digital Therapeutics for Heart Health

Use Case:

Deployment of AI techniques to personalize lifestyle change programs for heart disease patients incorporating environmental exposure data (e.g. weather pollution) to improve adherence and outcomes.

Outcome:

• Technical feasibility confirmed with TRL6 reached;
• End-to-end pipeline established: data collection, environmental enrichment, AI method assessment;
• Personalized support for behavior change using semi-automated decision-making;
• Defined hybrid AI-manual intervention workflow to maximize patient success.

Ecosystem Support:

The StairwAI program facilitated the formulation of a data-driven feasibility study, offered access to mentoring (both technical and business), and provided a structure for agile iteration across technical prototyping and business model innovation.

AI Relevance:

This case embodies AI democratization for health-focused SMEs by:
Advancing user-personalized AI with limited datasets;

1. Showcasing explainable integration of environmental data into digital therapeutics;
2. Enabling hybrid AI-manual interventions to increase trust and adoption;
3. Promoting sustainable business experimentation in underserved medical segments.

Summary:

AMYGDALAHEALTH OU is an Estonian digital health SME focusing on heart disease prevention through personalized behavioral change. The team developed AMYGML, an AI-enhanced version of their mobile app that augments existing behavior change routines with data on environmental conditions, such as pollution or extreme temperatures. The aim is to offer users dynamically adapted lifestyle advice, which is both scientifically grounded and context-aware. A feasibility plan was implemented, combining data collection, preparation for ML pipelines, and benchmarking of suitable AI models for personalization. In parallel, a business experiment was designed to evaluate how AI integration affects engagement and revenue potential, including randomized tests with customer segments. The company plans to continue this trajectory by refining its AI-enhanced intervention model and expanding testing with real users. This success story illustrates how AI can be ethically and effectively embedded into preventive health strategies—without compromising user experience or clinical rigor.