This paper presents an interpretation of the latest international technical report, IEC TR 62926-2019 (Medical electrical system – Guidelines for safe integration and operation of adaptive external beam radiotherapy systems) for real-time adaptive radiotherapy. It outlines the background for the development of this report, analyzes general safety guidelines for adaptive radiotherapy systems, and discusses the key design elements required for the integration of such systems. Additionally, the paper reviews and summarizes two typical reference models for adaptive external beam radiotherapy systems. The aim is to enhance the understanding and implementation of this technical report and to support its potential adaptation into a national standardized guidance document in China.

In the context of the ongoing advancement of precision medical education, governance in medical education within teaching hospitals has gradually shifted from traditional experience-based management to data-supported management. However, current approaches still largely remain at the stages of educational evidence-chain construction, profiling analysis, and risk prediction, with relatively limited capacity to support intervention consequence simulation, governance strategy comparison, and system-level optimization. Against this background, the concept of the medical education governance digital twin (MEGDT) is proposed. By integrating digital twin theory, data governance practices in teaching hospitals, and international frontier cases, this paper discusses the conceptual connotation, implementation framework, and governance value of MEGDT. The potential value of MEGDT lies not only in enhancing the dynamic perception, simulation, and feedback capabilities of medical education governance, but also in providing decision support for teaching hospitals to achieve a better balance among educational effectiveness, resource input, organizational efficiency, and educational equity. At present, MEGDT remains at the stage of conceptual proposal and pathway exploration. Future work should prioritize minimum viable prototype studies in scenarios such as postgraduate medical education, while also addressing data quality, model credibility, privacy and security, and cost-effectiveness balance.

Key Words: precision medical education; digital twin; medical education governance; evidence-informed decision-making

Standardized rehabilitation training for patients with coronary artery disease following coronary revascularization surgery is a core therapeutic strategy to improve postoperative cardiac function, boost exercise tolerance and optimize long-term prognosis. However, mainstream cardiac rehabilitation methods currently suffer from three prominent limitations: monotonous training scenarios, poorly individualized regimen titration, and poor adherence to home-based exercise regimens. In recent years, alongside iterative advances and technical maturation in artificial intelligence (AI), the Zhongshan-specific metaverse platform featuring virtual-physical integration has offered novel insights into overcoming bottlenecks in cardiac rehabilitation training. Leveraging metaverse-based immersive interactive technology and wearable dynamic electrocardiogram (ECG) monitoring technology, this study aims to develop an integrated intervention system tailored to standardized cardiac rehabilitation for coronary artery disease patients post coronary revascularization. It further defines the system’s implementation workflow, risk control protocols and clinical eligibility criteria. The research adopts four methodological approaches: system architecture development, scenario-based modular design, alignment with evidence-based clinical guidelines, and hierarchical competency-based teaching modeling. The postoperative cardiac rehabilitation framework consists of three core functional modules: virtual reality (VR) immersive exercise training, wearable real-time dynamic ECG data acquisition, and AI-powered ECG risk early warning. Rehabilitation phases are stratified according to patients’ postoperative functional capacity. On this basis, we established individualized immersive rehabilitation scenarios, graded exercise intervention thresholds, a tiered judgment system for ECG abnormalities, and standardized emergency response workflows. Meanwhile, to meet clinical teaching demands in cardiac rehabilitation departments, this study explores a sustainable specialty teaching transformation model supported by the Fudan Zhongshan Huisheng Intelligent Education platform. The findings will provide a comprehensive theoretical framework, technical roadmap and empirical evidence for the large-scale clinical translation of this innovative cardiac rehabilitation protocol.

Key Words: metaverse; immersive exercise training; cardiac rehabilitation; medical teaching transformation

Pain is a subjective and complex symptom, making its prediction, management, and treatment a significant chal lenge in clinical research. To address these challenges, the search for reliable and objective pain biomarkers has become a focal point in pain studies. Electroencephalography (EEG), a non-invasive clinical tool, has emerged as the most widely used method for assessing brain regions associated with pain due to its temporal resolution, ac curacy, and comprehensive nature. Multichannel EEG is now a primary technique in the study of pain biomarkers. This review discusses the current status and future prospects of EEG biomarkers in pain research, synthesizing evidence on the potential of EEG recordings as reliable biomarkers for pain perception. This will contribute to es tablishing a more solid foundation for the prediction, diagnosis, and intervention of pain in future research and management.

Medical ultrasound imaging is an integral part of preoperative diagnosis, lesion screening and ultrasound-guided interventional surgeries. Image segmentation techniques can enhance the identification of lesions and separate them from complex backgrounds, aiding physicians in both quantitative and qualitative analyses. Ultrasound image segmentation algorithms are primarily categorized into two types: traditional non-semantic segmentation and deep learning-based semantic segmentation, each with distinct advantages and drawbacks. This paper delves into these segmentation principles, elucidating their relevance in the realm of ultrasound image segmentation, and offers an overview of current research trends. Our goal is to provide guidance for physicians and researchers in selecting the most suitable segmentation algorithm that tailors to their specific requirements.