Citation

Sampa, Masuda Begum and Abdul Aziz, Nor Hidayati and Rahman, Md. Siddikur and Ab. Aziz, Nor Azlina and Besar, Rosli and Ghazali, Anith Khairunnisa (2026) Reinforcement learning for medical image analysis: a systematic review of algorithms, engineering challenges, and clinical deployment. Computer Assisted Surgery, 31 (1). ISSN 2469-9322

Text Reinforcement learning for medical image analysis_ a systematic review of algorithms, engineering ch.pdf - Published Version Restricted to Repository staff only Download (2MB)

Abstract

Reinforcement learning (R) has emerged as a powerful artificial intelligence paradigm in medical image analysis, excelling in complex decision-making tasks. his systematic review synthesizes the applications of R across diverse imaging domains—including landmark detection, image segmentation, lesion identification, disease diagnosis, and image registration—by analyzing 20 peer-reviewed studies published between 2019 and 2023. R methods are categorized into classical and deep reinforcement learning (DR) approaches, focusing on their performance, integration with other machine learning models, and clinical utility. Deep Q-Networks (DQN) demonstrated strong performance in anatomical landmark detection and cardiovascular risk estimation, while Proximal Policy Optimization (PPO) and Advantage Actor-ritic (A2) achieved optimal policy learning for vessel tracking. Policy gradient methods such as RNFOR, win-Delayed Deep Deterministic Policy Gradient (D3), and Soft Actor-ritic (SA) were successfully applied to breast lesion detection, white-matter connectivity analysis, and vertebral segmentation.Monte arlo learning, meta-R, and A3 methods proved effective for adaptive questioning, image quality evaluation, and multimodal image registration. o consolidate these findings, we propose a unified Reinforcement earning Medical maging (RM) framework encompassing four core components: state representation, policy optimization, reward formulation, and environment modeling. his framework enhances sequential agent learning, stabilizes navigation, and generalizes across imaging modalities and tasks. Key challenges remain, including optimizing task-specific policies, integrating anatomical contexts, addressing data scarcity, and improving interpretability. his review highlights R’s potential to enhance accuracy, adaptability, and efficiency in medical image analysis, providing valuable guidance for researchers and clinicians applying R in real-world healthcare settings.

Item Type:	Article
Uncontrolled Keywords:	Medical image analysis
Subjects:	R Medicine > R Medicine (General) > R858-859.7 Computer applications to medicine. Medical informatics
Divisions:	Faculty of Engineering and Technology (FET)
Depositing User:	Ms Rosnani Abd Wahab
Date Deposited:	11 Feb 2026 01:35
Last Modified:	11 Feb 2026 01:35
URII:	http://shdl.mmu.edu.my/id/eprint/15337

Downloads

Edit (login required)