Application of Human-Machine Collaborative Reinforcement Learning in Dynamic Management of Oral Anti-Obesity Drug Therapy

Read full text Summarise with VIRAT
Authors: Wang Yawen, Wang Yingli, Chen Jiaqi
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

In order to deal with the variability of treatment and decreasing adherence in AOMs, a patient-side app-based human-machine collaborative adaptive management framework is proposed in this study. Just like the direct "automatic prescription dosage adjustment", reinforcement learning (RL) is positioned as a decision support tool for generating candidate recommendations and optimizing follow-up schedules. The final protocols are established by clinical guidelines and medical confirmation, ensuring practical implementation and regulatory acceptability.We built a state space on the basis of application logs and wearable data (including weight change rate, postprandial blood sugar fluctuation, sleep efficiency, nausea score, step trends, etc.). The Action Space was defined as "Lifestyle Fine-Tuning Recommendation Combination + Follow Reminder Intensity + Recommendation to Recommend Medical Reassessment," which included safety restrictions (Risk Threshold and Contraindication Rules) .In the 10-week trial cohort, this strategy resulted in an additional 2.3% reduction in body weight relative to fixed follow-up and static recommendations, and a 21% reduction in the discontinuation rate for adverse events. This study demonstrates that RL is better suited as "continuous management and recommendation orchestrator" for patient-facingplatforms, forming a closed-loop system with front-end precision screening: screening → initiation → follow-up → adjustment → rescreening.

Fields of Study & Subjects

Mobile Health and mHealth Applications Artificial Intelligence in Healthcare and Education Pharmacology and Obesity Treatment

Publication details

Journal / source
Zenodo (CERN European Organization for Nuclear Research)
Publication date
25 March 2026
Language
EN
Document type
Preprint