Magnetic resonance-guided linear accelerator (MR-Linac) emerged as a novel technology that transformed the mode and process of radiation therapy (hereinafter referred to as “radiotherapy”). By virtue of its capabilities in clear soft tissue visualization, real-time dynamic tracking and on-line adaptivity, this system enabled more accurate contouring of target volumes and organs at risk, increased the radiation dose delivered to tumors while reducing the dose received by normal tissues, and thus achieved the goal of real-time and precise radiotherapy. This paper elaborated on the development history of radiotherapy, technical characteristics, clinical applications and key review points of MR-Linac, and made an outlook on its future development. This study could enhance the understanding of relevant industry practitioners, R&D personnel and regulators on magnetic resonance-guided radiotherapy products, provide a reference for the development of magnetic resonance-guided radiotherapy technology in China as well as the formulation of relevant technical guidelines and regulatory methods, and further promote the high-quality development of high-end and large medical equipment.

Given the lack of standardization and inconsistent operating procedures in the current clinical use of CT equipment, which were prone to cause equipment failures and affect diagnosis and treatment efficiency as well as patient safety, this study focused on the standardization of clinical use of CT equipment, relied on the special project of the National Key Research and Development Program, and formulated the group standard T/ZHYL 017-2025 “Specification for Clinical Use in X-ray Computed Tomography (CT) Equipment”. Based on relevant laws, regulations and industry standards, this specification mainly covered clinical environment configuration, examination preparation, use settings and maintenance support, clarified specific requirements such as machine room layout, scanning parameter optimization, CT tube protection, quality inspection and fault diagnosis, and adapted to the needs of different clinical scenarios. The promotion and implementation of this specification could improve the safety, effectiveness and economy of the clinical application of CT equipment, standardize the examination process, enhance the quality of medical services, and had important significance for ensuring patient safety and promoting the development of smart medical care.

As a Class II non-active surgical instrument, the sterile disposable retractor is widely used for retracting tissues or skin to assist in exposing the surgical field. At present, the registration of such products in China is active; however, the lack of unified standards and evaluation guidelines has led to inconsistent evaluation criteria. Based on the current medical device regulations and relevant standards, combined with product characteristics and registration practice, this paper systematically summarized the key review points and common problems of such products in the sections of regulatory information, summary documentation, research data (performance, biocompatibility, sterilization, validity period and packaging), risk analysis, clinical evaluation, product technical requirements, as well as instructions for use and labeling. Several controversial issues were also discussed. This study aimed to provide guidance for medical device registrants to systematically prepare application dossiers, and to offer references for technical review institutions to unify evaluation criteria and improve review quality.

To promote the coordinated development of in vitro diagnostics (IVD) between China and ASEAN, this paper conducted a systematic comparative analysis on the registration and submission requirements for IVD products in China and ASEAN. With the continuous advancement of the Belt and Road Initiative, economic and trade cooperation between China and ASEAN in the fields of pharmaceuticals and medical devices had been increasingly strengthened. This paper compared the differences in the regulatory systems of IVD between China and ASEAN from the aspects of regulatory regulations, product risk classification, registration dossiers, and market access pathways, and analyzed the implementation status in key regional countries. Results showed that China and ASEAN shared the same regulatory goal of ensuring product safety and effectiveness, but existed significant differences in implementation details, including classification rules, dossier requirements, reference country schemes, and registration procedures. Strengthening cooperation between China and ASEAN in the unification of IVD registration standards, regulatory mutual recognition, and regulatory coordination would help improve the efficiency of cross-border market access, ensure product quality and safety, and boost the high-quality development of the regional medical and health industry.

Objective To design a tracheal intubation system based on oropharyngeal airway guidance and magnetic positioning navigation technology, to achieve accurate and rapid tracheal intubation while reducing soft tissue injury. Methods The magnetically positioned and navigated tracheal intubation system was composed of a flexible oropharyngeal airway, a flexible magnetic positioning device, a flexible magnetic guiding strip, and a navigation magnet. The flexible oropharyngeal airway was first inserted into the oral cavity and adjusted to fully expose the epiglottis and glottis. The flexible magnetic guiding strip was inserted along the oropharyngeal airway, so that the guiding strip was prevented from bending or changing direction from the incisors to the glottic opening during blind intubation. The tip position of the flexible magnetic guiding strip was tracked and monitored by the magnetic positioning module in vitro. When the magnetic guiding strip reached the glottic opening area, the magnet at the tip of the guiding strip was attracted and navigated by the external navigation magnet to bend directionally and enter the trachea. Then the tracheal tube was inserted along the magnetic guiding strip, and the guiding strip was removed. The flexible oropharyngeal airway was retained in the oral cavity as a bite block and for catheter fixation. A tracheal blind intubation experiment was performed on an experimental simulation platform with 50 non-medical participants, each performing 10 operations. Learning time, first-attempt success rate, and intubation duration were recorded and analyzed to evaluate the application advantages of the system in tracheal intubation. Results With the assistance of the oropharyngeal airway guidance combined with magnetic positioning navigation system, blind tracheal intubation was performed on a human simulator. The mean first-attempt success rate reached 94%; 86% of operators achieved an intubation success rate of ≥90%, and 98% achieved ≥80%, showing stable and reliable overall performance. All operators completed intubation within 37 s. Among the total 470 successful intubations, 89.1% were finished within 30 s and 51.4% within 25 s. Conclusion The tracheal intubation system based on oropharyngeal airway guidance and magnetic positioning navigation shows obvious advantages. It effectively improves the first-attempt success rate of tracheal intubation in difficult airways, shortens operation time, and has low dependence on operators with good universality and application prospects. It can provide useful enlightenment for the development of intelligent medical robots for tracheal intubation and is worthy of further research and clinical application.

Objective To address three major problems in hysteroscopic image lesion detection: high computational complexity, low recall rate for tiny lesions (<5 mm), and difficulty in deployment on primary medical devices, to propose a lightweight realtime detection model. Methods A multi-scale feature pyramid (MSFP) architecture was constructed to enhance small lesion feature extraction via cross-layer feature fusion. A depthwise separable self-attention (DSSA) module was designed, and feature correlation was decomposed into intra-channel self-correlation and pointwise convolution to reduce the number of parameters. A dynamic region attention (DRA) mechanism was introduced to dynamically adjust the weights of 4×4 grid regions based on the lesion distribution probability matrix, so as to improve localization robustness. Gradient-sensitive pruning and 8-bit integer quantization were combined to achieve efficient model compression. Results On the HS-CMU dataset (3385 hysteroscopic images), the model achieved an average precision (AP@0.5) of 82.1% at an intersection over union threshold of 0.5, with only 2.1 M parameters and an inference speed of 48 FPS. The recall rate for tiny lesions smaller than 5 mm was improved to 76.3%. Under interference of Gaussian noise (σ=0.1) and motion blur (kernel=15), the recall rate fluctuation was less than 3.5%, which was significantly superior to mainstream models. Conclusion The model achieves a balance between accuracy and efficiency through the collaborative design of MSFP-DSSADRA. It provides a low-cost, high-robustness solution for real-time hysteroscopic lesion localization in primary medical institutions, assists clinicians in improving disease detection rate, and possesses important value for clinical promotion and application.

Objective To solve the problems of complex fault types and low efficiency of manual diagnosis for switching power supplies in medical equipment, to propose an intelligent fault diagnosis method based on improved convolutional neural network (CNN). Methods A total of 30000 time-series signal samples of 4 categories were collected, and a multi-channel dataset was constructed. Batch normalization, L2 regularization, squeeze-and-excitation (SE) attention mechanism and AdamW optimizer were introduced into CNN to improve feature extraction efficiency and robustness. Performance was evaluated using accuracy, recall rate and F1 score. Noise disturbance and data missing experiments were carried out to verify model stability. Results The improved CNN achieved an accuracy of 99.2%, a recall rate of 98.9% and an F1 score of 99.0% on the test set. Among optimizers compared, AdamW performed best, with model convergence within 100 epochs and a single-sample inference time of 5.2 ms, significantly outperforming SGD and Adam. In ablation experiments, the accuracy of traditional CNN was 96.2%, which rose to 97.5% and 98.7% after adding regularization and SE attention mechanism successively, and the full improved strategy achieved the optimal performance. Compared with SVM, random forest and traditional CNN, the accuracy was improved by 7.7%, 6.1% and 2.0% respectively. In robustness tests, the accuracy was 95.4% at noise variance σ2 =0.10 and still reached 89.8% under strong noise with σ2 =0.20. The accuracy remained above 96.2% within 20% data missing and 93.1% at 30% data missing. Conclusion The proposed method presents high accuracy, strong robustness and deployment feasibility. It provides technical support for intelligent equipment maintenance.

Objective To improve the prediction accuracy of the risk of residual urinary calculi after flexible ureteroscopic lithotripsy and to reduce the postoperative re-intervention rate. Methods A risk prediction model for residual calculi after lithotripsy was proposed based on optimized extreme gradient boosting (XGBoost). The model combined an improved sparrow search algorithm to optimize the XGBoost model. Circle chaotic mapping was introduced for population initialization, nonlinear decreasing inertia weight and Levy flight strategy were adopted to improve the quality of hyperparameter search. Comparative experiments were carried out using the CICIDS 2018 dataset and postoperative follow-up data of patients with urinary calculi who underwent flexible ureteroscopy assisted by an intelligent pressure control system at No.1 People’s Hospital of Dali City from January 2023 to December 2024. Results Experimental results showed that the optimized XGBoost model achieved an area under the curve (AUC) of 0.94 on both the training set and the test set, with a classification accuracy of 92.49%. In normal flow scenarios, the AUC reached as high as 0.96, and both the F1 score and recall rate reached 0.91, which were significantly superior to those of the comparative models. Conclusion The model can stably predict the risk of residual calculi and provides reliable support for preoperative risk assessment and personalized treatment.

Objective To construct an intelligent heart rate monitoring method based on remote photoplethysmography (rPPG) and 3D convolutional neural network (CNN), to significantly improve detection accuracy through the collaborative optimization of deep learning and signal processing. Methods The combination of rPPG and 3D CNN was proposed in this study, and singular spectrum analysis and adaptive filtering algorithms were introduced for further optimization. In the effectiveness evaluation experiments, the selected evaluation indicators included the mean absolute error, root mean square error, and Pearson correlation coefficient between the predicted values and the ground truth. Meanwhile, the Bland-Altman plots and confidence intervals were adopted for result analysis. Results Ablation experiments revealed that the proposed algorithm achieved a mean absolute error of 0.46 and a root mean square error of 0.91 in the PURE dataset. In the UBFC-rPPG dataset, the mean absolute error was only 1.46, which was lower than that of the single CNN and CNN-rPPG models. The confidence interval of the proposed method was [-2.5, 2.5], and the predicted curve showed high consistency with the ground truth heart rate curve. Conclusion The vital sign detection method based on rPPG and 3D CNN accurately estimates heart rate and presents significant advantages in long-term vital signs monitoring.

Objective To address the problems of low protection efficiency, insufficient security and high cross-device deployment cost for medical device data, to propose a novel data protection model for improving the security and economy of the entire data life cycle. Methods A two-layer consensus network was adopted to optimize the Byzantine fault-tolerant algorithm, and efficient consensus decision-making was realized through hierarchical collaboration. A data encryption layer was constructed combined with the advanced encryption standard algorithm, forming an integrated model of consensus and encryption. Based on the MDPCD and MedOdyssey datasets, experiments were carried out to verify the model by comparing it with traditional encryption models, single consensus models and mainstream blockchain schemes in terms of attack resistance, communication efficiency and device compatibility. Results The experimental results showed that the optimized Byzantine fault-tolerant algorithm had only 231 communication times at a maximum node size of 55, and an average latency as low as 986 ms at 150 nodes. The medical device data protection model built on this algorithm achieved the highest score of 10 in three core security indicators: Byzantine attack resistance, dynamic key update and differential privacy support, all significantly superior to all control models. Meanwhile, high compatibility was achieved in all types of medical devices with low, medium and high computing power. The total deployment cost was only (95±4.7) US dollars, which was lower than that of the control models, with a statistically significant difference (P<0.001). Conclusion The proposed algorithmic model significantly improves the efficiency and security of data protection for various medical devices, and also exerts a positive effect on reducing the cost of data protection.

Objective To investigate the economic impact of surgical costs and usage trends of flow diverter (FD) devices made of different materials in the treatment of intracranial aneurysms. Methods Clinical application data of FD with different materials from Xuanwu Hospital of Capital Medical University between January 2022 and March 2025 were selected to analyze the overall usage trend. The usage quantities and total costs of two types of FD from 2022 to 2025 were statistically analyzed, and the Wilcoxon rank-sum test was used to compare the monthly differences in the usage of the two materials across different years. Additionally, the Wilcoxon rank-sum test was employed to compare the total medical costs and total consumable quantities for patients using the two materials in each quarter from 2022 to 2025, aiming to comprehensively evaluate the cost-effectiveness of FD with different materials. Results The differences in the quantity and cost of FD usage between the two materials from 2022 to 2025 were statistically significant (P<0.05). However, there was no statistically significant difference in the total quantity of consumables used by patients with either material during the same period (P>0.05). The total medical costs for patients using the two materials showed a statistically significant difference in 2022 and 2023 (P<0.05), but no significant difference in 2024 and 2025 (P>0.05). Conclusion Against the backdrop of current clinical practice and volume-based procurement policies, the utilization rate of cobalt-chromium alloy stents is significantly higher than that of nickel-titanium alloy stents. However, there is no significant difference in terms of direct medical economics between the two types of stents. Cobalt-chromium alloy drug-eluting stents have gained broader clinical application and maintained a growing trend of utilization due to their superior performance.

Objective To evaluate the accuracy of dose delivery for carbon ion radiotherapy in uveal melanoma based on off-line positron emission tomography (PET)/CT radionuclide imaging, and to verify its reliability in clinical treatment planning. Methods Forty-nine patients with uveal melanoma who underwent carbon ion radiotherapy were enrolled. Off-line PET imaging was performed immediately after treatment to obtain the activity distribution of β+ radionuclides. By comparing the PET activity distribution with the dose distribution predicted by the treatment planning system (TPS), the dose conformity index (DCI) and PET conformity index (PCI) were calculated, and three-dimensional dose delivery characteristics were evaluated. The accuracy of dose delivery was quantified using the dose delivery index (DDI), defined as the ratio of PCI to DCI. Results In the 95% dose region, DCI was 59.64% ± 14.61%, PCI was 58.13% ± 14.43%, and DDI was 99.52% ± 8.63%. In the 90% dose region, DCI was 42.96% ± 11.22%, PCI was 42.01% ± 11.08%, and DDI was 97.50% ± 11.80%. Differences in DCI and PCI between the 95% and 90% dose regions were statistically significant (P<0.001), whereas no significant difference in DDI was observed between the two groups (P=0.645>0.05). Tumor volume showed no significant effect on DDI (all P>0.05). Transit time had a significant influence on dose delivery conformity (P=0.047<0.05). Target location (left/right eye) differed in some indices, suggesting that right-eye targets may present certain challenges in image registration or dose delivery. Conclusion Off-line PET radionuclide imaging serves as an effective tool for dose verification in carbon ion radiotherapy of uveal melanoma, with high accuracy especially in the assessment of target dose deposition. This study provides an important reference for online dose monitoring of carbon ion therapy and the correlation between dose and PET activity.

Objective To Design and develop an asset inventory management system based on industrial Internet of Things technology to achieve rapid inventory of hospital assets and improve the completeness of asset records. Methods Using MyEclipse6.5 software as the development tool, the system was designed in Java language and adopts a B/S architecture. It used MySQL database as the backend management software for the system design. The core functional modules included login management, user management, master data management, inventory management, and data processing. Results Black-box testing was conducted, and the system achieved a 100% pass rate in terms of performance. Before and after the application of the system, the average time required for inventory counting was significantly reduced from 120.28 h to 16.30 h, with a significant difference (P<0.05). Additionally, the average accuracy rate of inventory counting increased from 85.46% to 98.52%, with a significant difference (P<0.05). Conclusion The study not only addresses the core pain point of “discrepancy between records and reality” in hospital asset management, but also, through the deep integration of a standardized master data system and Internet of Things technology, provides digital tools for refined asset operation and management. It is of great significance in helping medical institutions optimize resource allocation, activate state-owned assets, and promote the digital transformation of hospitals.

Objective To address the issues of high construction costs, large data analysis volumes, and low efficiency in existing medical record management systems, an intelligent medical record information management system is designed that considers both cost and operational risks, in order to achieve automation in medical record information management. Methods Based on the cloud platform, optimized the medical record information management system from three aspects: data analysis, storage access, and functional design of the medical record system. This included building a MySQL database structure, designing a data interaction framework and automatic query mode, and constructing a Naive Bayes disease diagnosis auxiliary model. The designed intelligent case information management system is examined and analyzed with the help of accuracy, recall, evaluation metrics, throughput, response time, and cost. Results Plain Bayes outperforms other comparative methods in terms of accuracy and recall >0.85 for disease classification with inference delay values<2 ms. The throughput response time exhibited by the research-designed cloud platform server is short, and the batch statistics time for data access of case information is 1.8 s, and the risk of data loss is only 0.001%. Conclusion The intelligent medical record information management system designed for research can effectively meet the needs of medical information management, improve application efficiency, and reduce operation and maintenance costs and risks.

Objective To explore the solution of the electronic form system based on the PaaS mode, optimize the full-process management of low-value consumables in PIVAS and improve the level of informatization. Methods Utilizing the PaaS-based BaiShu low-code development platform, with front-end interaction, business logic, data services, and application output as the foundational architecture, and incorporating features such as user-friendly operation, automatic loading, expiration warnings, inventory verification, and real-time inventory statistics, an electronic form management system was designed to cover core processes including acceptance of incoming goods, consumable requisition, consumable return, and consumable inventory. The application effectiveness was evaluated through data comparison and questionnaire surveys. Results A three-month period before and after system implementation was selected as the control and study groups, respectively. The average account-material match rate increased from (89.03±4.89)% to (95.53±1.95)% (P<0.05). The average registration time decreased from (193.88 ± 20.91) s to (108.04 ± 10.94) s (P<0.05); 95.45% of the respondents recognized the system functions and were willing to continue using them. 86.36% believed that the data accuracy had significantly improved, and 81.81% thought it could effectively enhance work efficiency. However, there is still room for improvement in interface design and compatibility with mobile devices. Conclusion The construction of an electronic form management system based on the low-code development platform of the PaaS model can effectively solve the shortcomings of traditional manual management, significantly improve the management efficiency and data reliability of low-value consumables in PIVAS, and provide medical institutions with a low-cost and easy-to-implement informatization path. In the future, we can explore the expansion of application scenarios and attach importance to data security.

Objective To analyze the cost control effects of four medical insurance supervision models for medical consumables, clarify the advantages, disadvantages and risk points of different models, and provide a reference for medical insurance administrative departments and medical institutions to improve the medical insurance supervision system for the use of medical consumables. Methods The monitoring data of inpatient medical consumables expenses in the Second Affiliated Hospital of Wenzhou Medical University from January 1, 2024 to January 31, 2025 were collected, with weekly data extraction to form 58 evenly spaced time periods. Taking the switching time of different management models as the interruption points, a multi-stage interrupted time series analysis was adopted to compare the medical insurance violation rate and self-funded violation rate under various management models. Results Under the four models, the medical insurance violation rates were 2.36%, 1.25%, 0.13% and 0.01% respectively, and the self-funded violation rates were 0.30%, 4.32%, 3.63% and 0.35% respectively. The pop-up reminder and approval model significantly reduced the medical insurance violation rate of inpatients, but significantly increased the self-funded violation rate. From the perspectives of reducing medical insurance and self-funded violation rates and cutting down the workload of discrimination, the intelligent discrimination and reminder model was significantly superior to the other three management models. Conclusion The combination of information transformation and manual expense monitoring in medical institutions can realize the efficient management of medical insurance violations in inpatient medical consumables charging, yet the fundamental improvement relies on the improvement of rules by medical insurance administrative departments and the establishment of a full-process management mechanism in medical institutions. The safe and stable operation of the medical insurance fund depends on the high-pressure supervision of medical insurance administrative departments, while it is imperative to guard against the medical insurance moral hazard caused by the hierarchical transfer of medical insurance violation risks.

Objective To conduct multi-dimensional analysis and evaluation of the usage data of high-value medical consumables in the whole hospital based on diagnosis related groups (DRG), so as to provide data support for establishing the usage specifications of high-value medical consumables. Methods Multi-information systems were integrated to collect full-process data of high-value consumables, with Nanjing Drum Tower Hospital, The Affiliated Hospital of Nanjing University Medical School as the research subject. Multi-dimensional analyses were conducted from perspectives including hospital-wide trends, consumable utilization by departments and physicians, and consumable distribution for the same surgical procedures. Based on procedure data corresponding to DRG disease groups and combined with expert demonstration, specifications for consumable use were established, and a hierarchical early warning mechanism was constructed. Results From 2021 to 2023, the annual compound growth rate of consumables usage in the whole hospital was 6.06%, with high-value medical consumables accounting for over 75% of the total; the top 5 departments contributed 54.56% of the total consumables usage. Taking the department of thoracic surgery and video-assisted thoracoscopic lobectomy as examples, key consumables (such as staplers and hemostatic anti-adhesion materials) and physicians with abnormal per-case costs were identified. After establishing consumable usage specifications and the hierarchical early warning mechanism, in the surgical procedure for right upper lobe malignant tumors, the overuse rate of staplers decreased from 24.65% to 19.94%, and that of hemostatic anti-adhesion materials decreased from 22.99% to 9.66%. The quantity, cost of relevant consumables, and total cost of high-value medical consumables all decreased significantly, with statistically significant differences (P<0.05). Conclusion The multidimensional analysis and hierarchical early warning system for high-value medical consumables based on multi-system linkage can realize precise supervision and standardized management of consumables usage, effectively reduce consumables costs, and provide a feasible tool for the refined operation of hospitals under DRG payment reform, which has high clinical application and promotion value. In the future, it is necessary to further improve the dynamic adjustment mechanism of early warning indicators to enhance management precision.

Objective To conduct an in-depth analysis of procurement, operation and maintenance costs as well as usage data between domestic and imported large-scale medical devices using municipal hospitals in Beijing as an example, and to provide references and suggestions for equipment demonstration and the sound development of domestic medical devices. Methods Data on in-service CT, magnetic resonance imaging (MRI), and X-ray equipment [including digital subtraction angiography (DSA), digital radiography (DR), C-arm, O-arm, and G-arm] from 22 municipal hospitals in Beijing were collected and investigated. The rank-sum test was used to analyze cost differences in three aspects: the comparison of procurement prices between domestic and imported products, the comparison of procurement prices between domestic-brand products and localized products of foreign brands, and the comparison of equipment maintenance costs. The actual usage volume and procurement proportion of domestic and imported products were compared to evaluate differences in equipment availability. Results Procurement prices of domestic products were significantly lower than those of imported products for six types of equipment, including 64-slice (inclusive) to 128-slice (exclusive) CT, 1.5 T and below MRI, above 1.5 T MRI, DSA, DR and C-arm (P<0.05), with a reduction range of 12.97%-50.80%. Among five categories [CT below 64-slice, 64-slice (inclusive) to 128-slice (exclusive) CT, 1.5 T and below MRI, DR and C-arm], only C-arm showed no significant difference in procurement prices between domestic-brand products and localized products of foreign brands (P>0.05); procurement prices of domestic-brand products were significantly lower than those of localized foreign-brand products for the other categories (P<0.05), with a reduction range of 19.68%-31.91%. For four types of equipment [64-slice (inclusive) to 128-slice (exclusive) CT, 1.5 T and below MRI, DR and C-arm], the full-service maintenance rates of domestic and imported products after the warranty period were basically the same (P>0.05). The procurement proportion of domestic equipment surpassed that of imported equipment for the first time in 2020 and remained stable in the following five years. No significant difference was found in the average usage volume of CT and MRI equipment between domestic and imported products (P>0.05). Conclusion Overall, the procurement price and maintenance cost of domestic products are lower than those of imported products, and the usability and market acceptance of domestic products are not inferior to those of imported products. Medical institutions should strengthen demonstration and research to reduce equipment procurement and operation costs on the premise of meeting clinical needs.

Objective To systematically clarify the current research status of biomedical materials applied in the field of rehabilitation and accurately predict the future development trends of this field. Methods Based on the literatures retrieved from the Web of Science Collection, the research literatures on biomedical materials in rehabilitation from the database establishment to December 2024 were analyzed. Methods including Bradford’s Law, Lotka’s Law and cited reference publication year spectroscopy were adopted to conduct visual analysis on the information of journals, authors, cited literatures and other aspects. Results A total of 1965 valid literatures were included. The annual number of published papers and the average annual citation frequency in this field both showed an upward trend, with Biomaterials and other 9 journals being the core ones. Murphy WL was identified as the core author; 80.705% of the authors published only one relevant paper, and all core authors had formed their own research teams. The University of Wisconsin System in the United States and Beihang University in China were the leading publishing institutions at home and abroad, all of which had formed cooperative clusters. China and the United States maintained close research cooperation, while Germany had the highest degree of international cooperation; there was a positive correlation between the scale of scientific research output and academic influence in this field. “Bone regeneration” “spinal cord injury” and “3D printing” were the current research hotspots, and “rehabilitation” and “artificial intelligence” were emerging research themes. The research from 2017 to 2024 focused on “3D printing” “hydrogel” and other directions. Conclusion This study clarifies the core research subjects and development context of biomedical materials in the field of rehabilitation. The field is showing an overall fluctuating upward development trend, and the intellectualization of biomedical materials, the integration of 3D printing technology with biomedical materials, and the cross-disciplinary combination of materials and genetics will become the future development trends, which will provide technical support for the precise development of rehabilitation medicine.

Objective To explore the evaluation of musculoskeletal function adjacent to the dorsal root ganglion (DRG) and its correlation using magnetic resonance imaging (MRI) combined with musculoskeletal ultrasound (MUS) in patients with zosterassociated pain (ZAP). Methods Patients with ZAP admitted to affiliated hospital of Hebei university from June 2020 to July 2023 were enrolled and divided into the postherpetic neuralgia group (n=45) and herpes zoster neuralgia group (n=64). General data, MRI and MUS parameters, Oswestry Disability Index (ODI), Visual Analog Scale (VAS) were compared between the two groups. Multiple linear regression analysis was performed to assess the correlation between MRI parameters and musculoskeletal function indicators adjacent to DRG. Multivariate logistic regression analysis was used to identify independent risk factors for postherpetic neuralgia. Restricted cubic spline models were used to analyze the nonlinear dose-response relationship of MRI parameters, MUS parameters, and musculoskeletal function adjacent to DRG with postherpetic neuralgia. Receiver operating characteristic (ROC) curves and calibration curves were used to evaluate the predictive accuracy of the model. Results There were significant differences in hypertension history and smoking between the two groups (P<0.05). Patients with postherpetic neuralgia had significantly higher surface volume ratio, maximum 2D diameter column, median of grayscale values of DRG, large area high gray-level emphasis of DRG, MUS score, numerical rating scale (NRS) score, ODI and VAS than patients with herpes zoster neuralgia (all P<0.001). Multivariate logistic regression showed that hypertension, smoking, surface volume ratio, maximum 2D diameter column, median of grayscale values of DRG, short-run low gray-level emphasis, large area high gray-level emphasis, MUS score and NRS score were independent factors for postherpetic neuralgia (all P<0.05). Restricted cubic spline analysis showed a nonlinear dose-response relationship between MRI parameters, MUS parameters, musculoskeletal function adjacent to DRG and the risk of postherpetic neuralgia. The ROC curve yielded an area under the curve of 0.859 (95% confidence interval: 0.813-0.902), and the Brier score of the calibration curve was 0.087, indicating high predictive performance. Conclusion MRI combined with MUS imaging features can evaluate musculoskeletal function adjacent to DRG in patients with ZAP. MRI and MUS parameters are correlated with musculoskeletal function parameters adjacent to DRG, and the model shows high discrimination and accuracy.

Cardiovascular disease poses a major threat to the health of Chinese residents, and the prevalence of hypertension is extremely high. Traditional blood pressure measurement methods are limited by invasiveness and intermittency, which make it difficult to meet the demand for continuous monitoring under natural conditions. Wearable continuous blood pressure measurement devices based on photoplethysmography (PPG) have become a research hotspot due to their non-invasiveness and portability. However, the measurement accuracy of such devices is restricted by multiple factors including physiology, environment and algorithms. In addition, the standardization system and clinical validation mechanism are still immature, and the international regulatory thresholds are increasingly elevated, resulting in numerous challenges for industrial development. This paper reviewed the technical principles and classifications of such devices, analyzed the factors influencing their measurement accuracy, elaborated on the current status of the product market, clinical research and regulation, and interprets relevant international standards. It also points out the challenges in PPG technological development and prospects future directions, aims to provide references for the research and development and clinical validation of this technology, promote its standardization and mature application, and support the prevention and control of cardiovascular diseases and health management of residents in China.

As a major challenge in the global public health field, cardiac arrest is characterized by high mortality and disability. The core mechanism underlying these outcomes lies in cerebral ischemia-reperfusion injury caused by prolonged “no-flow” state of brain tissue and low perfusion during the resuscitation period. Real-time acquisition of cerebral oxygenation parameters during resuscitation is crucial for improving prognosis. As a core noninvasive and continuous indicator for monitoring cerebral oxygenation, regional cerebral oxygen saturation (rSO2) enables bedside dynamic evaluation via near-infrared spectroscopy and has been widely used in critical care medicine, rehabilitation medicine and surgical disciplines. In cardiopulmonary resuscitation (CPR), monitoring of rSO2 can reflect the balance between cerebral oxygen supply and demand in real time, providing critical evidence for optimizing compression quality, predicting return of spontaneous circulation, evaluating cerebral blood flow metabolism, guiding targeted temperature management, judging neurological prognosis, and making decisions on extracorporeal CPR, thus demonstrating unique clinical application value. This article reviewed the technical advances and multi-scenario applications of rSO2 monitoring in the field of CPR in recent years, aims to provide evidence-based references for optimizing resuscitation strategies and improving patient prognosis.

With the continuous development of mobile intelligent technology, its application in the home-based rehabilitation of stroke patients with limb motor dysfunction has gradually become a research hotspot. Currently, a variety of technical means, such as virtual reality, exoskeleton robots, wearable devices, smart home and Internet of Things systems, mobile rehabilitation management applications, and remote video guidance, play an important role in improving patients’ rehabilitation adherence, reducing medical costs, and enhancing their disease self-management capabilities. However, the application of this technology faces multiple challenges, including complex device operation, data security and privacy protection, high device costs, and strong reliance on networks. This paper reviewed the application forms, effects, and shortcomings of mobile intelligent technology in the home-based rehabilitation of stroke patients with motor dysfunction, aiming to provide a theoretical basis and practical reference for constructing an intelligent, individualized, and sustainable home-based rehabilitation model for stroke patients in the future.

Osteoporosis has a high incidence rate and has become the fourth most common chronic disease worldwide, imposing a heavy burden on healthcare systems. At present, the main therapeutic approaches for osteoporosis include pharmacotherapy, exercise therapy and rehabilitation therapy, yet these conventional treatments are plagued by such problems as poor treatment adherence and limited therapeutic efficacy. As an emerging form of medical intervention, digital therapeutics exerts a pivotal role in the treatment, prevention and health education of osteoporosis by integrating application software with technologies including virtual reality, artificial intelligence and wearable devices. Despite significant progress achieved in relevant research, there is a lack of systematic collation of the findings. This paper reviewed digital therapeutics in terms of its general overview, application forms, clinical effects in osteoporosis as well as the existing problems and corresponding solutions, aiming to provide a reference for the further application of digital therapeutics in the management of osteoporosis.

Against the backdrop of the continuous development of the smart healthcare system, artificial intelligence has been widely applied in the field of medical care. As a core department for surgical treatment and rescue, the operating room is characterized by high risks, and its risk management exerts a direct impact on hospital operation and patients’ life safety. Clinically, nearly 50% of adverse events in hospitals occur in the operating room, more than half of which are preventable. At present, the core nursing risks in the operating room focus on five aspects: handover of surgical patients, surgical safety verification, intraoperative pressure injury, perioperative hypothermia, and pathological specimen management. Existing intervention measures mostly center on improving medical staff’s risk awareness and applying quality improvement tools. However, relevant research on artificial intelligence in the field of operating room risk management has rarely been reported, and the exploration of its intelligent application remains inadequate. This paper summarized the current application status of artificial intelligence in this field at home and abroad, conducted an analysis on the above five high-risk factors, clarified the current research status and limitations, and discussed and prospects the development direction of intelligent technologies for operating room risk management in China. It is intended to provide a reference for the further development of artificial intelligence technology in this field, help improve the efficiency of operating room risk management, safeguard the surgical safety of patients, and promote the construction of intelligent operating rooms.

Ischemic stroke (IS) is a major cerebrovascular disease leading to disability and death worldwide. Abnormal cerebral hemodynamics is its core pathological process, and accurate monitoring of cerebral microcirculatory blood flow is crucial for mechanistic research and optimized diagnosis and treatment. Conventional cerebral blood flow monitoring techniques have defects such as insufficient spatiotemporal resolution, reliance on exogenous contrast agents, and inability to conduct long-term real-time monitoring, which can hardly meet the research and clinical needs of IS. As an emerging optical imaging technology, Laser speckle contrast imaging (LSCI) shows unique value in IS research by virtue of its non-invasiveness, high spatiotemporal resolution and real-time blood flow monitoring capability. This paper reviewed the technical principle of laser speckle imaging, its application in the establishment and validation of IS animal models, clinical translation potential, technical limitations and future development directions. It covered the latest research progress of LSCI in cerebral hemodynamic monitoring, collateral circulation assessment and therapeutic intervention effect evaluation, with special focus on its innovative applications in the delineation of acute ischemic penumbra, reperfusion therapy monitoring and neurovascular coupling research. By comparing the advantages and disadvantages with traditional imaging techniques, aims to provide a new technical perspective for basic research and clinical diagnosis and treatment of IS, and discuss the future trend of multimodal imaging integration.

Echocardiography serves as a core non-invasive tool for the diagnosis of cardiovascular diseases. However, its dynamic imaging characteristics give rise to difficulties in the standardization of image acquisition, and the analytical process heavily relies on operator experience. These factors lead to significant discrepancies in diagnoses among different physicians, restricting diagnostic efficiency and objectivity. In recent years, breakthrough advances in artificial intelligence (AI), particularly deep learning (DL) technologies, have provided new opportunities for optimizing the echocardiography workflow. This paper reviewed the latest research progress and application value of DL-based AI techniques in echocardiographic image analysis, focused on the applications of AI in core links including image segmentation, cardiac function evaluation, and cardiovascular disease diagnosis. Meanwhile, it thoroughly analyzed the current challenges and prospects future development directions, aims to provide a reference for clinicians to understand the application value of AI tools and achieve more efficient and accurate diagnosis of cardiac diseases.

Objective To systematically analyze faults of digestive endoscopy equipment based on real clinical data and to propose targeted preventive maintenance strategies, so as to improve equipment efficiency, reduce maintenance costs and ensure medical safety. Methods A retrospective analysis was conducted. A total of 183 cases of endoscopy fault data from the Digestive Endoscopy Center of the First Hospital of Lanzhou University between January 2021 and December 2024 were selected as research objects. The fault rate, fault phenomena, causes, locations and maintenance levels were comprehensively analyzed with chisquare test. Results The overall fault rate of digestive endoscopes was 0.16%, and the distribution difference was statistically significant (P<0.001). Highfrequency fault phenomena included angle problems (0.11%) and component wear/damage (0.11%). Among the causes, wear and aging accounted for the highest proportion (38.96%). Highincidence fault locations were concentrated in the insertion tube (35.65%) and the operation section (33.18%), and the distribution difference of locations was statistically significant (P<0.05). Maintenance level D accounted for the highest proportion (62.30%), and the distribution difference of maintenance levels was statistically significant (P<0.001). Highfrequency faults of digestive endoscopes mainly occurred in those with service life of 1-3 years and 7-9 years. Conclusion Based on the analysis of real endoscopy fault and maintenance data, this study puts forward a comprehensive optimization scheme covering standardized training, multilevel maintenance system, targeted maintenance for highfrequency faults, information management, improved endoscopy management system and extended service life, which provides a theoretical basis for reducing the clinical endoscopy fault rate.

Objective To conduct an in-depth analysis of the fault causes of the “system over-temperature” alarm for the Maquet CARDIOSAVE Hybrid intra-aortic balloon pump (IABP), to summarize maintenance experience, and to put forward improved measures for equipment management, so as to enhance the operational efficiency and reliability of the equipment. Methods The maintenance processes of 6 cases of “system over-temperature” alarm faults of this model of IABP were retrospectively analyzed. The fault mechanisms were analyzed in detail combined with the system structure and working flow of the equipment. Meanwhile, the usage status of the existing IABP equipment in the hospital was statistically collected, including the information of the using departments, cumulative service time, service life and fault frequency. Results Six types of fault causes and corresponding maintenance solutions were summarized, and it was found that the fault rate of over-temperature alarms caused by problems of negative pressure and positive pressure filters was relatively high. The usage frequency of IABP equipment varied significantly among different departments, and the overall equipment allocation of the hospital still had room for optimization. A positive correlation was observed between the service life of the equipment and the number of faults, which was consistent with the law of the equipment failure curve. Conclusion Equipment manufacturers should optimize equipment design, expand the monitoring scope and add more monitoring indicators. Medical institutions should implement scientific equipment management, including formulating maintenance cycles and exploring a shared equipment model, to reduce the fault rate and improve the comprehensive benefit of IABP equipment.