The consequences induced by the use of general anesthetics and better options of the drugs in children and infants have been topics of controversy due to the concerns about their potential impact on cognitive function. To address these concerns and ensure the safety of pediatric general anesthesia, this paper reviews existing basic and clinical studies that have investigated the mechanisms of general anesthetics on pediatric cognitive function. In this paper, the basic research on neurotoxicity of general anesthetics and the clinical research on the effects of general anesthesia drugs on cognitive function in children in recent years were analyzed and summarized. Although some of the clinical studies have suggested that general anesthesia in children or infants may cause neurocognitive damage and a series of behavioral complications, the results of the retrospective studies need to be viewed with caution as they may lack effective control for relevant factors that could have impacts in the perioperative period. It remains uncertain whether general anesthetics affect the cognitive function of pediatric patients. Further research is needed to establish clear evidence-based recommendations for clinical prevention and treatment measures to minimize the potential risks associated with the use of general anesthetics.

Histology laboratory time is often consumed by basic orientation that students could complete before class. We therefore evaluated a blended model that moved part of this preparation online and reserved the face-to-face session for observation, questioning, and inquiry. The study used a quasi-experimental design and included 252 undergraduates from six intact clinical medicine classes in the 2025–2026 academic year. Three classes (n=126) used the blended model and three classes (n=126) received conventional laboratory instruction. Before each laboratory session, students in the intervention group used Rain Classroom to review microlectures and related materials and to complete self-assessments. Laboratory time was then organized around targeted questions, tiered inquiry tasks, group discussion, and instructor feedback. Outcomes included a common end-of-course microscopic structure identification examination, an anonymous questionnaire, and semi-structured interviews. No significant baseline differences were found between groups (P>0.05). The intervention group scored 86.37±5.48 on the final examination, compared with 80.72±6.31 in the control group (P<0.001). Among the 118 valid questionnaire respondents, 92.37% considered the blended model more effective than conventional laboratory teaching, 89.83% reported that preclass learning clarified the laboratory objectives, and 83.90% agreed that tiered inquiry improved their understanding of structure-function relationships. Interview comments echoed these findings, especially with regard to preparation, repeated review of morphological material, and participation during laboratory work. Within the setting studied, reorganizing preparation and laboratory activities in this way was associated with better examination performance and a favorable learning experience.

Clinical care requires iterative evidence acquisition, revision of working hypotheses, and action under incomplete information, changing patient states, and finite resources. Yet medical artificial intelligence (AI) is still evaluated largely through recognition, prediction, content generation, or static question answering. The central scientific question is whether model capability can be translated into safe, effective, and accountable longitudinal clinical action. A medical agent is not merely a large language model (LLM), but a sociotechnical system integrating models, knowledge, data, memory, tools, workflows, human oversight, and governance. Drawing on representative studies, authoritative reviews, reporting guidance, and regulatory documents available through July 2026, this Review proposes a six-module closed-loop architecture—perception, knowledge, reasoning, planning, action, and feedback—clarifies boundaries with related AI systems, and links clinical use cases to evidence and governance requirements. Retrieval-augmented generation, tool use, reflection, and multi-agent collaboration have improved performance in selected controlled tasks. However, static medical question-answering scores are weak proxies for interactive clinical performance, and recent benchmarks show only modest accuracy gains despite substantial increases in token use and latency. A small number of randomized trials and real-world workflow studies suggest potential benefits for documentation, trial prescreening, and bounded decision processes, but do not support unsupervised autonomous diagnosis or treatment. We therefore propose four risk-calibrated autonomy levels, a six-stage evidence pathway, and lifecycle governance. Clinical value should be judged by the net benefit of the clinician–agent–workflow system. The near-term goal is not clinician-independent autonomy, but clinician-led, patient-engaged, evidence-constrained, and accountable human–agent care.

Key Words: medical AI agent; large language model; multi-agent system; clinical translation; human–AI collaboration; trustworthy artificial intelligence; lifecycle governance

With the advancement of metaverse technology and the successful commercialization of head-mounted displays, the application prospects of the metaverse in medical education are increasingly prominent. This article systematically reviews the latest advancements in the application of the metaverse in otolaryngology and head and neck surgery, and proposes ideas for specialty construction. VR surgical simulators, 3D printing combined with VR technology, can significantly improve trainee assessment scores and operational confidence, optimizing hand-eye coordination and spatial perception abilities. AR microscope systems enable contactless interaction through gesture recognition, enhancing surgical efficiency. AR navigation platforms demonstrate real-time 3D visualization advantages in craniofacial surgery, and VR and 360° video training effectively improve the anterior nares packing skills of junior doctors. VR-assisted teaching for peritonsillar abscess drainage significantly enhances medical students' structured clinical examination scores, operational confidence, and engagement. Integrating 3D virtual models and image overlay technology in transoral robotic surgery provides practical value for image-guided surgery. VR technology can effectively alleviate pain and anxiety in patients undergoing otolaryngological surgery, improve satisfaction, and serve as a non-pharmacological intervention to reduce opioid use. VR combined with wearable devices can reduce postoperative opioid consumption, and VR vestibular rehabilitation training can enhance treatment compliance and quality of life for patients with otolith disorders. VR has become an innovative component of ENT digital teaching, demonstrating equivalent or even superior effects compared to traditional methods in anatomy, physiology, and pathology teaching. Based on these advancements, this article proposes the idea of relying on the "Fudan Zhongshan Huisheng Zhiyu" metaverse platform to build a specialized teaching system for otolaryngology: constructing high-fidelity 3D digital twin models and a structured case resource library, forming a tiered training path covering institutional education, post-graduation education, and continuing medical education, promoting the dissemination and sharing of high-quality teaching resources, and driving the innovative development of otolaryngology medical education in China.

Key Words: full-cycle medical education; training and management; Fudan Zhongshan Huisheng Zhiyu; metaverse platform; virtual reality

Radiofrequency ablation (RFA) represents a convenient, minimally invasive, and cost-effective approach for the treatment of small liver cancers measuring less than 3 cm in diameter. Nonetheless, the existing RFA techniques encounter challenges in precisely controlling the extent of ablation and the risk of overheating, which can lead to damage on the surrounding tissues. The ability to control the ablation area also plays a crucial role in the success of RFA procedures. To address these issues, we introduce a novel method that utilizes tumor boundary temperature monitoring to achieve precise control over the RFA process. Through the utilization of COMSOL Multiphysics simulation software, the proposed method was verified by comprehensive simulation and modeling and its efficacy in achieving regional control was testified. Subsequently, ex vivo experiments were conducted employing a custom-designed cooling RFA instrument. The experimental results demonstrated that the RFA controlled by boundary temperature yielded an ablation area with a diameter approaching 30 mm. In comparison to the standard spherical solidification zone (characterized by both the ellipticity and regularity indices of 1), the cooling RFA discoloration zone, under boundary temperature control, exhibited a maximum deviation of 7% in the ellipticity index and 17% in the regularity index. The average ellipticity index was determined to be 0.96, while the average regularity index was 0.86. Collectively, these findings underscore the capability of the proposed temperature-controlled cooling RFA method to attain precise control over the ablation area, contributing to comprehensive ablation of tumor tissue within the target region.

Specialty education in neurology faces structural bottlenecks including insufficient training in low-frequency, high-risk clinical scenarios, unequal opportunities for performing key procedures, difficulty in replicating ethically sensitive communication situations, and a weak cross-stage evaluation system. Based on the “Huisheng Intelligent Education” metaverse platform of Zhongshan Hospital, Fudan University, this paper proposes a conceptual framework for building a neurology-specific metaverse platform targeting full-cycle medical education, training, and management. Using extended reality (XR) and artificial intelligence (AI) as the technological foundation, and focusing on core teaching scenarios such as neurological examination, emergency decision-making in stroke, status epilepticus management, long-term care of Parkinson‘s disease, and breaking bad news, the platform constructs four major modules: a neurology knowledge ecosystem, immersive virtual teaching scenarios, a specialized teaching intelligence system, and a full-cycle teaching management hub. The platform supports cross-campus synchronous teaching and process tracking, aiming to establish a new paradigm of specialty education characterized by “scenario-based training — formative evaluation — personalized guidance — governance closed loop”. Taking the “acute ischemic stroke thrombolysis decision-making metaverse teaching unit” as an example, this paper demonstrates the feasible approach of the platform in scenario design, task chain organization, and technical implementation. This paper provides a replicable engineering framework and design reference for the systematic construction of metaverse teaching in neurology, with future effectiveness validation to be conducted based on actual operational data.

Key Words: neurology; metaverse in medicine; full-cycle medical education; immersive training; thrombolysis decision-making; competency evaluation

Objective: To design and validate a portable multi-degree-of-freedom electric needle holder to improve operational flexibility and precision in laparoscopic surgery, while reducing the operator's workload. The design optimizes the transmission system to enable precise control of surgical instruments, addressing the challenges of complex surgical environments. Methods: A multi-degree-of-freedom transmission structure, driven by steel wires, was designed. It integrates motors and angle sensors to achieve multi-directional rotation and force control for the grasping forceps. The experimental section includes tests for maximum gripping force, maximum extraction force, and a porcine suturing experiment. The performance of the electric needle holder was compared with traditional needle holders in simulated laparoscopic surgery to verify its effectiveness. Results: Experimental results show that the maximum gripping force of the electric needle holder at various angles reached 29.6 N, significantly higher than that of traditional needle holders. The maximum extraction force for needles of different sizes is approximately 12.9 N, effectively preventing needle slippage. Additionally, the porcine suturing experiment demonstrated that the electric needle holder significantly reduced operation time and improved suturing precision in complex surgical settings. Conclusion: The portable multi-degree-of-freedom electric needle holder enhances operational flexibility and safety in laparoscopic surgery by increasing the rotational degrees of freedom and precise motor control. Its superior gripping force and operational efficiency suggest strong clinical potential, offering significant improvements over traditional surgical instruments.