![[2602.19298] ALPACA: A Reinforcement Learning Environment for Medication Repurposing and Treatment Optimization in Alzheimer's Disease](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.19298] ALPACA: A Reinforcement Learning Environment for Medication Repurposing and Treatment Optimization in Alzheimer's Disease
Summary
The paper presents ALPACA, a reinforcement learning environment designed for optimizing medication repurposing and treatment strategies in Alzheimer's disease, leveraging existing therapies through a novel simulation model.
Why It Matters
ALPACA addresses the challenges of evaluating personalized treatments in Alzheimer's disease, which is often hindered by lengthy clinical trials and patient variability. By providing an open-source platform for simulating treatment strategies, it has the potential to accelerate research and improve patient outcomes.
Key Takeaways
- ALPACA is an open-source, Gym-compatible RL environment for Alzheimer's treatment optimization.
- The Continuous Action-conditioned State Transitions (CAST) model generates realistic treatment trajectories.
- RL policies trained in ALPACA outperform traditional treatment baselines.
- The platform emphasizes interpretability by relying on clinically relevant patient features.
- ALPACA serves as a reusable testbed for individualized treatment decision-making.
Computer Science > Artificial Intelligence arXiv:2602.19298 (cs) [Submitted on 22 Feb 2026] Title:ALPACA: A Reinforcement Learning Environment for Medication Repurposing and Treatment Optimization in Alzheimer's Disease Authors:Nolan Brady, Tom Yeh View a PDF of the paper titled ALPACA: A Reinforcement Learning Environment for Medication Repurposing and Treatment Optimization in Alzheimer's Disease, by Nolan Brady and Tom Yeh View PDF HTML (experimental) Abstract:Evaluating personalized, sequential treatment strategies for Alzheimer's disease (AD) using clinical trials is often impractical due to long disease horizons and substantial inter-patient heterogeneity. To address these constraints, we present the Alzheimer's Learning Platform for Adaptive Care Agents (ALPACA), an open-source, Gym-compatible reinforcement learning (RL) environment for systematically exploring personalized treatment strategies using existing therapies. ALPACA is powered by the Continuous Action-conditioned State Transitions (CAST) model trained on longitudinal trajectories from the Alzheimer's Disease Neuroimaging Initiative (ADNI), enabling medication-conditioned simulation of disease progression under alternative treatment decisions. We show that CAST autoregressively generates realistic medication-conditioned trajectories and that RL policies trained in ALPACA outperform no-treatment and behavior-cloned clinician baselines on memory-related outcomes. Interpretability analyses further indicated t...
