Abstract:The rapid increase in the percentage of chronic disease patients along with the recent pandemic pose immediate threats on healthcare expenditure and elevate causes of death. This calls for transforming healthcare systems away from one-on-one patient treatment into intelligent health systems, to improve services, access and scalability, while reducing costs. Reinforcement Learning (RL) has witnessed an intrinsic breakthrough in solving a variety of complex problems for diverse applications and services. Thus, we conduct in this paper a comprehensive survey of the recent models and techniques of RL that have been developed/used for supporting Intelligent-healthcare (I-health) systems. This paper can guide the readers to deeply understand the state-of-the-art regarding the use of RL in the context of I-health. Specifically, we first present an overview for the I-health systems challenges, architecture, and how RL can benefit these systems. We then review the background and mathematical modeling of different RL, Deep RL (DRL), and multi-agent RL models. After that, we provide a deep literature review for the applications of RL in I-health systems. In particular, three main areas have been tackled, i.e., edge intelligence, smart core network, and dynamic treatment regimes. Finally, we highlight emerging challenges and outline future research directions in driving the future success of RL in I-health systems, which opens the door for exploring some interesting and unsolved problems.

What problem does this paper attempt to address?

This paper attempts to solve several key problems faced by intelligent medical systems (I - Health systems). Specifically: 1. **Meeting the challenges brought by the growth of chronic disease patients and the aging population**: With the increase in the number of chronic disease patients and the acceleration of global population aging, the demand for medical services has risen sharply. This not only increases the burden on the medical system but also places higher requirements on medical costs and service quality. 2. **Improving the intelligence level of medical services**: Traditional medical systems mainly rely on the one - to - one patient treatment model, which is difficult to meet the needs of modern society for efficient, convenient, and scalable medical services. Therefore, it is necessary to transform medical services into intelligent health systems to improve services, enhance accessibility and scalability, and reduce costs at the same time. 3. **Using reinforcement learning to optimize the performance of intelligent medical systems**: In order to achieve the above - mentioned goals, the paper explores how to optimize all aspects of intelligent medical systems by introducing reinforcement learning (Reinforcement Learning, RL), especially deep reinforcement learning (Deep Reinforcement Learning, DRL). RL has significant advantages in dealing with sequential decision - making problems in complex environments and can help intelligent medical systems better respond to dynamically changing environments, process massive amounts of data, and make optimal decisions. 4. **Solving the technical challenges of intelligent medical systems**: - **High - dynamic environment**: The intelligent medical system faces diverse and irregular data generation and collection methods, and the network environment is also highly dynamic, making it difficult for traditional methods to respond effectively. - **A large number of potential users**: With the increase in chronic disease patients and the elderly population, the demand for medical services has surged, making it extremely difficult to provide accurate and personalized clinical decisions. - **Distributed and unbalanced data**: The locally generated data from different patients is unevenly distributed, which may lead to biases and inaccurate predictions in machine - learning algorithms. - **Limited computing and communication resources**: The large number of patient participation makes the availability of computing and communication resources a major challenge, especially on edge devices. 5. **Constructing the architecture of intelligent medical systems**: The paper proposes a three - layer architecture including the perception layer, the edge intelligence layer, and the intelligent core network layer, aiming to solve the above - mentioned challenges. This architecture extends from data generation/collection in the physical layer to the control and management layer, ensuring the effective processing, storage, and transmission of data. In summary, the main purpose of this paper is to explore how to use reinforcement learning technology to construct and optimize intelligent medical systems to meet the many challenges currently faced in the medical field and to point out the direction for future research and development.

Reinforcement Learning for Intelligent Healthcare Systems: A Comprehensive Survey

Reinforcement Learning in Healthcare: A Survey

Reinforcement learning for intelligent healthcare applications: A survey

Reinforcement Learning Algorithms and Applications in Healthcare and Robotics: A Comprehensive and Systematic Review

Data Quality, Bias, and Strategic Challenges in Reinforcement Learning for Healthcare: A Survey

Challenges for Reinforcement Learning in Healthcare

Reinforcement Learning for Clinical Decision Support in Critical Care: Comprehensive Review

Reinforcement Learning in Modern Biostatistics: Constructing Optimal Adaptive Interventions

Adaptive Multi-Agent Deep Reinforcement Learning for Timely Healthcare Interventions

A Primer on Reinforcement Learning in Medicine for Clinicians

Deep Reinforcement Learning for Clinical Decision Support: A Brief Survey

Bias in Reinforcement Learning: A Review in Healthcare Applications

A Review of Reinforcement Learning for Natural Language Processing, and Applications in Healthcare

Does Reinforcement Learning Improve Outcomes for Critically Ill Patients? A Systematic Review and Level-of-Readiness Assessment

Recent Reinforcement Learning and Blockchain Based Security Solutions for Internet of Things: Survey

The roots of Jewish sobriety. Alcoholism in the first century C.E

Is Deep Reinforcement Learning Ready for Practical Applications in Healthcare? A Sensitivity Analysis of Duel-DDQN for Hemodynamic Management in Sepsis Patients

A survey on medical image analysis using deep reinforcement learning approach

Optimal Treatment Strategies for Critical Patients with Deep Reinforcement Learning

Reinforcement Learning in Education: A Literature Review

Reinforcement learning for intensive care medicine: actionable clinical insights from novel approaches to reward shaping and off-policy model evaluation