Objective In view of the problems existing in clinical practical training in traditional Chinese medicine (TCM) education and teaching, such as low learning interest, relatively fragmented time, and rather monotonous learning and memorization methods, it is proposed to integrate information technology means and build a traditional Chinese medicine clinical classic practical training platform (TCM-CCTP) that combines education with entertainment. Methods Utilize the SpringBoot development framework as the business processing technology, present it through a visual WeChat mini-program application, adopt MySQL + Redis for data storage. Meanwhile, by analyzing the combat requirements of the Traditional Chinese Medicine Clinical Classic Practical Training Platform (TCM-CCTP), a message synchronization and algorithm matching scheme combining WebSocket technology and the Redis storage framework is proposed to achieve the core business functions of the platform.. Results After the implementation of the platform, an analysis of chapter - based test data reveals a notable enhancement in users' depth and breadth of understanding regarding traditional Chinese medicine (TCM) clinical knowledge. The difference is statistically significant (t = 2.368, P = 0.012). In a quantitative assessment of the platform - based learning model, which has a total score of 5 points, the satisfaction score for its promotion and application reached 4.95. This indicates that the platform exerts a positive influence on learning outcomes, facilitating a more effective acquisition of TCM clinical knowledge. Conclusion The construction and implementation of the Traditional Chinese Medicine Clinical Classic Practical Training Platform (TCM-CCTP) have increased the participation in clinical practical training, strengthened students' ways of mastering knowledge, integrated students' time, effectively reduced the learning and economic costs, and promoted the active development of practical training work on classic traditional Chinese medicine.

Objective To build explainable models based on clinical knowledge and compare them with traditional black-box models to further improve the effectiveness of Mayo endoscopic score of ulcerative colitis. Methods A total of 2251 endoscopic images from the Endoscopy Center of the First Affiliated Hospital of Soochow University were collected as the training set and internal validation set. The black box model was constructed by categorizing the images into groups 0, 1, 2, and 3 according to the Mayo endoscopic score. At the same time, all images were labeled according to five features, including ulceration (presence or absence), spontaneous hemorrhage (presence or absence), erythema (without, visible, obvious), vascular texture (normal, fuzzy, disappearing), and mucosal friability (normal, mild, easily friable), to create a sub-featured model, which was fused to construct an interpretable model. In addition, endoscopic images from the Endoscopy Center of Changshu Hospital affiliated with Soochow University were used for external validation. In the external validation set, the performance of the explainable model was compared with the black-box model by calculating the accuracy, Matthew correlation coefficient (MCC), and Cohen's Kappa with two endoscopists of different years of experience. Finally, the Grad-CAM method was used to highlight the regions on which the model's reasoning was based. Results The accuracies of the four explainable models based on MobileNet, ResNet, Xception, and EfficientNet in the external validation set were 0.765, 0.800, 0.830, and 0.885, respectively, which were better than those of the corresponding traditional black-box models, 0.665, 0.705, 0.775, and 0.815, with the explainable model based on EfficientNet performing the best, outperforming low seniority physicians (0.805) and high seniority physicians (0.870). Conclusion In this study, four explainable computer vision models were constructed and externally validated by collecting endoscopic images of ulcerative colitis, and the optimal model is selected, suggesting that the explainable model performs better than the traditional deep-learning black-box model in the Mayo endoscopic score of ulcerative colitis. The explainable model has good application value in the future endoscopic diagnosis of ulcerative colitis.

Objective Sterile cold spray is the first approved sterile medical device of this type in China. During its development, optimizations were made to the refrigerant and valve actuator, and innovations were introduced to the manufacturing process, achieving product sterilization. The spray offers a localized, immediate analgesic effect for superficial tissues. From December 2023 to June 2024, we conducted an observation study on the effectiveness and safety of sterile cold spray in alleviating venipuncture pain and the associated nervousness and fear in first-time blood and plasma donors. Methods Using an open, randomized controlled, multi-center approach, a trial of sterile cold spray to reduce pain and donor nervousness and fear was carried out on 480 first-time non-remunerated blood donors and 2,296 plasma donors at the blood and plasma stations, respectively; the donors were randomly divided into an observation group (n=240) and a control group (n=240), and similarly, the plasma donors were divided into an observation group (n=1146), a control group (n=1150), and a control group (n=1150), and the plasma donors were divided into an observation group (n=1146), and a control group (n=1150). , control group (n=1150); the observation group was given a sterile cold spray for pain intervention before venipuncture by spraying a sterile cold spray at the puncture site for 4-10 seconds and the control group was operated as usual and each group was administered the Visual Analogue Scale (VAS) and Blood Donor Anxiety Scale (BDAS) questionnaires. A telephone callback survey on blood donation intention was also conducted for first-time donors; a telephone callback survey on satisfaction was conducted for first-time plasma donors. Results There were no significant differences between the groups in terms of gender, education, age, height, weight, and BMI; pain scores in the blood donor observation group: 2.75±1.85, and anxiety scores: 14.97±2.39, and pain scores in the control group: 3.85±1.72, and anxiety scores: 16.76±2.62; and pain scores in the plasma donor observation group: 2.67±1.87, and anxiety scales: 15.13±2.20, pain score in the control group: 4.02±1.88, anxiety score: 16.90±2.67; the differences between the two groups were statistically significant (P< 0.001). In the telephone interview on willingness to donate blood, the proportion of the observation group willing to donate blood again was 72.1%, higher than that of the control group, which was 48.0% (P< 0.01); and in the telephone interview on plasma donors' satisfaction with pain management, the observation group was satisfied with 86.31%, higher than that of the control group, which was 61.05% (P< 0.01). Conclusion For first-time blood donors or plasma donors, the use of aseptic cold spray intervention during venipuncture can reduce their pain, relieve their anxiety, and help increase their willingness to donate again.

Objective To evaluate the performance differences between domestic and imported ultrasound diagnostic equipment in a clinical setting, and to provide a basis for the clinical selection and use of ultrasound diagnostic equipment, this paper will conduct a comparative analysis of the sensitivity and center frequency of ultrasound probes. Methods The study is based on the pulse-echo method and employs the FirstCall and a self-developed ultrasound probe testing system. Using a comparative research method, a random sample of 54 ultrasound probes in clinical use was selected. The crystal sensitivity and center frequency, which are key indicators affecting probe performance, were chosen, and an evaluation standard was established to analyze and evaluate the test results of the ultrasound probes.Results  The study found that there was no significant difference in sensitivity between domestic and imported abdominal probes, while the sensitivity of domestic superficial probes was superior to that of imported brands, especially when the usage period was 5 years or more. Additionally, there was no significant difference in the frequency deviation between domestic and imported probes. Conclusion The performance of the ultrasound probes tested in this study showed no significant difference between imported and domestic brands. Additionally, this study provides scientific methods and data support for evaluating the performance of domestic ultrasound probes, and it lays a research foundation for the clinical effectiveness and consistency evaluation of ultrasound diagnostic equipment performance testing, which helps establish a more comprehensive re-evaluation mechanism for ultrasound equipment in clinical use.

Objective To explore the impact of “3444” management mode on optimizing resource allocation and enhancing service efficiency in the development of the integrated urban medical groups. Methods Developing the "3444" management model aims to guide management efforts in building integrated urban healthcare groups. Its core lies in establishing three major platforms, establishing four key systems, focusing on four major disease categories, and enhancing four core capabilities, thus forming a comprehensive closed-loop management process for various tasks within the integrated urban medical groups. Results Under the "3444" management model, the core hospital provides various professional training programs to primary healthcare institutions, promoting the continuing education and on-the-job training of grassroots medical personnel. Additionally, the core hospital dispatches responsible directors and consulting experts to primary healthcare institutions, cumulatively serving 2,317 patients with a total service duration of 1,123 hours. In the first three quarters of 2024, the number of referrals between the core hospital and primary healthcare institutions were 22,971, 24,429, and 23,448, respectively. Furthermore, the core hospital has implemented a remote electrocardiogram (ECG) project that has reached a wide range of primary healthcare institutions, with a total of 6,610 individuals screened as reported by these institutions. The service quality at primary healthcare institutions has been assessed, and the assessment scores significantly increased after implementing the "3444" management model (P<0.05). Conclusion Under the "3444" management model, the allocation of medical resources among members of the integrated urban healthcare group has been optimized, leading to improvements in the quality and efficiency of medical services and enhanced service capabilities of primary healthcare institutions. This management model plays a positive role in the development of integrated urban healthcare groups.

Objective Based on the established evaluation index system of medical imaging dynamic acquisition techniques, multidimensional comprehensive evaluation, as well as strength and weakness analysis, were conducted on various techniques. And the results will provide decision-making foundation for selection of dynamic acquisition techniques. Methods Three types of acquisition techniques were arranged in hospital environment, including interconnection based on log analysis, interconnection based on host image recognition and data collection based on manual operation by original equipment manufacturers. Meanwhile, according to the established index and weight of the evaluation system through systematical information retrieval and expert consultation in the early stage, results were then analyzed using comprehensive scoring method. Results Operation data of medical imaging equipment were collected in a certain period of time using the three acquisition techniques mentioned above. Multidimensional evaluation on these techniques was arranged regarding five primary indexes, fourteen secondary indexes and thirty-six tertiary indexes. Conclusion Acquisition technique based on log analysis gained a higher score and advantage over the other two in multiple dimensions. When it comes to evaluation results in terms of tertiary indexes, all of the three techniques have room for improvement and it is suggested to form complementary strength through advantage integration.

Objective PDCA cycle method was used to continuously improve the quality management of infant incubators in neonatology department, and improve the qualified rate and guarantee ability of infant incubators. Methods Relevant data were collected, problems and causes were analyzed by fishbone diagram and Plato. Combined with PDCA cycle management, improvement measures were implemented by improving medical equipment management system, refining maintenance content, strengthening personnel training and assessment, and improving supervision and management. The improvement effect was evaluated according to the overall qualified rate and fine detection parameters before and after the introduction of PDCA cycle method. Results After implementing the improvement measures, the qualified rate in infant incubator increased from 87.1% to 97.1%, and the difference was statistically significant (P < 0.05). At the same time, the deviation of temperature index and relative humidity in the test items is obviously reduced, and the deviation, uniformity and fluctuation of temperature index have significant changes (P < 0.05). The results show that PDCA cycle method has improved the quality management level and performance status of infant incubator. Conclusion Using PDCA circulation method to improve the qualified rate and performance of infant incubator can promote the standardization of quality management in infant incubator and provide reference for other medical equipment to carry out quality management.

Objective To investigate the application effect of the full-process medical technology appointment platform in the "daily clearance" of patient examinations in the radiology department. Methods This study used a retrospective analysis method to collect patient visit data from the imaging department of Yantaishan Hospital in Yantai before and after the introduction of a full-process medical technology reservation platform, covering the time periods from January to June 2022 and January to June 2023. The platform design uses Java, JavaScript and other development languages, based on SpringCloud microservice architecture and Vue front-end framework, to build a full-process management system that integrates reservation, queuing, inspection and reporting. By collecting key indicators such as patient waiting time, completion rate of examination on the same day, and patient satisfaction, we used statistical methods to compare and analyze the data of the two groups to evaluate the application effect of the full-process medical technology reservation platform. Results After the introduction of the full-process medical technology appointment platform, the patient waiting time in the radiology department of Yantaishan Hospital was significantly shortened, from an average of 59.6±15.8 minutes to 21.7±6.2 minutes. The daily examination completion rate increased significantly, jumping from 61.36±12.15% to 85.28±11.98%. Additionally, patient satisfaction improved markedly, with overall satisfaction rising from 5.52±1.9 to 8.47±1.2. The complaint rate of patients has significantly decreased. The complaint rate regarding patients' waiting time has dropped from 0.53 ± 0.06 (%) to 0.12 ± 0.01 (%), and the complaint rate concerning patients' satisfaction has decreased from 0.33 ± 0.05 (%) to 0.08 ± 0.01 (%).All these data results were statistically tested, and the statistical test results indicated that there were significant differences (P<0.05) before and after the introduction of the full-process medical technology appointment platform. Conclusion The full-process medical technology appointment platform has played an important role in the "daily clearance" of patient examinations in the radiology department, effectively improving medical service quality and patient experience.

To address the inefficiency and high error rates of manual inspection of three-phase information (batch number, production date, and expiry date) on pharmaceutical packaging in traditional production processes, this study designs an automated pharmaceutical information recognition system based on Optical Character Recognition (OCR). The system aims to enhance quality control and ensure regulatory compliance. It integrates innovative hardware configurations with improved software technologies. On the hardware side, a detection platform was designed to accommodate pharmaceutical packages of various specifications, leveraging photometric stereo technology to capture high-quality embossed character images. On the software side, the text detection and recognition algorithms were improved. The enhanced DBNet algorithm incorporates deformable convolution modules and feature pyramid enhancement modules to improve detection capabilities for complex text scenarios, while the text recognition module integrates deformable attention mechanisms to enhance the distinction of complex characters. Experimental results show that the system achieves an accuracy of 89.0% when processing 1200 low-quality pharmaceutical information images, an improvement of 3.7 percentage points compared to the traditional ABINet model. Additionally, in real production line testing, the system achieved a detection accuracy of 98%-100% for five types of pharmaceutical packaging, with an average detection time approximately one-quarter that of manual inspection. This system significantly improves the efficiency and accuracy of pharmaceutical three-phase information detection and recognition, overcoming the limitations of traditional methods and providing reliable technical support for quality regulation in pharmaceutical production.

Objective This study aims to establish a set of quality control evaluation indicators for tender procurement management in order to further standardize procurement processes, ensure procurement quality, and enhance hospital management efficiency. Methods Using the Delphi method, 16 experts were consulted to establish evaluation indicators. Meanwhile, the Analytic Hierarchy Process is used to calculate the weights of each evaluation indicator. Results After two rounds of expert consultation, the evaluation system indicators were finally determined, covering 5 primary indicators and 19 secondary indicators. All determined indicator weights have passed consistency tests. Conclusion The evaluation indicator system constructed in this study is scientific and reasonable, aiding hospitals in controlling the quality of tender procurement management, ensuring fairness, justice, and transparency in the procurement process, and thereby enhancing the overall management level of the hospital.

Objective To explore a strategy for aligning multi-view medical images with medical reports to optimize the quality of automatically generated medical imaging reports by deep learning models. Methods A generation model integrating multi-view features and attention mechanisms was designed. Firstly, it encodes the features of medical images taken from different angles and the pathological description features of reports based on a pre-trained model. Then, various attention mechanisms are utilized to complete the fusion calculation of the features. Finally, the decoder is used to translate the combined features into case reports. Results After multiple rounds of testing on the authoritative and publicly available chest IU X-Ray datasets, the model was on average 8.34%, 14.2%, 10.9%, 6.14%, 1.7%, and 5.5% higher than the previously proposed methods in six typical evaluation metrics, with a comprehensive performance improvement of 7.79%. Conclusion The model performs well in terms of the accuracy and fluency of the generated reports, validating that the integration strategy effectively captures the underlying relationships between images and reports, thereby enhancing the model's report generation capabilities.

Bone age assessment plays an important role in monitoring the growth and development of children. However, the application of traditional X-ray imaging techniques is often affected by background interference, which reduces the accuracy of the assessment results. In recent years, deep learning technology has been widely applied in the field of bone age image processing, demonstrating significant advantages. This paper reviews the progress of deep learning in addressing background redundancy in bone age imaging. It focuses on methods such as region of interest extraction, background segmentation techniques, and attention mechanisms. Research indicates that these methods can effectively eliminate redundant background information, enhancing the accuracy and efficiency of bone age assessments. Additionally, this paper discusses the limitations of existing technologies and future development directions, aiming to provide new research insights and practical guidance for the field of bone age assessment.

Objective To study and solve the consistency and standardization issues of technical requirements for medical device products after registration and listing. Methods From the perspective of registration and evaluation of medical device products, and through a questionnaire survey, the technical requirements of medical device products and the requirements for preparing change comparison tables were analyzed; Based on the WPS Office platform, develop a medical device product technical requirements and product basic information management system using VBA (Visual Basic for Application). Results Construct a product technical requirement content data framework with a single clause as the minimum unit, and apply this data framework to its information management system. Implement automatic standardization of medical device product technical formats, assist in writing product technical requirements and product registration information change comparison tables. Compared with conventional operating methods, the average preparation cycle of product technical requirements is shortened by 35%, the average preparation cycle of product technical requirement change comparison tables is shortened by 70%, and the number of errors is reduced by 15%. Conclusion The establishment of this management system and data framework in this article will provide a new tool for internal management of medical device product technology and efficient integration of various stages of the medical device lifecycle, such as inspection and evaluation.

Objective The purpose of this study is to optimize the operation arrangement and improve the utilization rate of operating room and patient satisfaction through the intelligent operation scheduling system. Methods Based on the data center, the system was developed by SQL Server 2016, Visual Studio 2019 and C#, and realized the automatic scheduling of B/S website in the mode of Vue+.Net Core. Results By leveraging a data mid-platform, this system has achieved the intellectualization and automation of surgical scheduling and personnel rostering. It has also incorporated an intelligent reminder function for scheduling results. Furthermore, it has established a Business Intelligence (BI) system specifically for operating rooms, significantly enhancing the efficiency, convenience, and safety of surgical services within hospitals.After the application of the system, the preoperative waiting time was reduced by 16%, the overtime time of nurses was reduced by 16%, the efficiency of operation scheduling was increased by 62%, the frequency of operation adjustment was reduced by 35%, the error rate of operation was reduced by 60%, the average daily operation volume was increased by 8.2%, the patient satisfaction was increased by 1.3%, and the operation accuracy rate was increased by 22.5%. The differences in each index were statistically significant (P<0.05). Conclusions The intelligent surgical scheduling system is a pivotal tool in modern hospital management. It optimizes surgical scheduling through intelligent and automated processes, thereby improving operating room utilization, minimizing surgical delays, and enhancing both patient satisfaction and medical quality.

Objective To propose a denoising method for positron emission tomography (PET) images based on adaptive hybrid filtering and to validate its denoising effectiveness at different acquisition times. Methods The study dynamically adjusts the filtering window size and selects appropriate filter parameters to adapt to varying noise levels. To evaluate the method's effectiveness, PET images from 33 brain tumor patients were used, with three different acquisition times (3 minutes, 5 minutes, and 7 minutes) as experimental samples. The denoising effect of the adaptive hybrid filter was assessed by comparing PSNR, SSIM, and other metrics with traditional filtering methods. The validation process included statistical analyses such as Friedman test and Cohen’s Kappa coefficient, with P-values (P < 0.05) used to test the significance of results. Results The experimental results showed that at a 3-minute acquisition time, the adaptive hybrid filter (PSNR: 60.9569 ± 2.0467, SSIM: 0.9960 ± 0.0015, MAE: 6.6901 ± 2.1756, MSE: 0.0571 ± 0.0217) exhibited significantly better denoising effects compared to traditional methods (P < 0.05). At 5 minutes, the adaptive hybrid filter (PSNR: 62.0394 ± 2.2481, SSIM: 0.9971 ± 0.0012, MAE: 4.7381 ± 1.7955, MSE: 0.0450 ± 0.0174) significantly outperformed other methods in terms of PSNR, MAE, and MSE (P < 0.05). At 7 minutes, the adaptive hybrid filter (PSNR: 62.7323 ± 2.2265, SSIM: 0.9976 ± 0.0009, MAE: 3.8339 ± 1.5475, MSE: 0.0384 ± 0.0149) showed significantly better MAE than other methods (P < 0.05). The visual scores of the adaptive hybrid filter at all three acquisition times (3min: 2.64 ± 0.50; 5min: 3.18 ± 0.50; 7min: 3.55 ± 0.46) were significantly higher than those of other methods (P < 0.05). Conclusion The adaptive hybrid filtering method can effectively reduce noise while preserving image edges and details, making it suitable for PET image denoising. It demonstrates superior denoising performance and robustness across different noise levels and acquisition times.

Objective This paper analyzes the current situation of Class B large medical equipment allocation in China and the level of coupling coordination with high-quality economic development, identifies obstacles, and provides a reference for the rational optimization of Class B large equipment allocation.Methods The entropy weight method and comprehensive evaluation method were used to comprehensively evaluate the configuration of Class B large medical equipment and high-quality economic development, and the coupling coordination degree model was used to measure the level of coupling coordination between them. The main obstacle factors affecting the coupling coordination were calculated using the obstacle function.Results At the end of the "Thirteenth Five-Year Plan" period, the stock of Class B medical equipment in China was 4,501 units, and the "Fourteenth Five-Year Plan" targets 3,528 units. The total amount of Class B medical equipment is insufficient and shows a pattern of East > Central > West. Among the 31 provinces, 8 are in a state of high-quality coordinated development, 19 are in a state of good coordinated development, and 4 are in a state of moderate coordinated development. The average coupling coordination degrees for the eastern, central, and western regions are 0.8291, 0.7861, and 0.7261, respectively. The number of Class B large medical devices, the ownership of Class B large medical devices per million population, the level of economic green development, and the level of economic coordinated development are the main obstacle factors that constrain the coupling coordination level of this binary composite system.Conclusion Class B large medical equipment still requires government planning improvement; administrative departments at all levels should play a leading role in science, rationally allocate Class B large medical equipment; adhering to the concept of green, shared, and coordinated development, according to regional resource endowments, promote the coordinated development of the two systems in a manner adapted to local conditions.

Objective To establish an evaluation system for the centralized management mode of medical equipment, providing reference for the evaluation of management system and optimization of management mode. Methods An expert group was established to construct a preliminary draft of the evaluation index system for the centralized management mode of medical equipment through literature review and expert consultation. After two rounds of Delphi expert inquiry, the final index system was determined, and the weights of each level of indicators were determined through weight investigation and analysis. Results A total of 12 experts were invited to participate in the evaluation system research. The effective response rates of the two rounds of inquiry questionnaires were both 100%, with authoritative coefficients of 0.88054 and 0.84303. The coefficient of variation ranges were 0.0952-0.3934 and 0.0805-0.2513, respectively, The Kendall coordination coefficients are 0.339 and 0.434, respectively, indicating a high degree of coordination. Finally, 3 primary indicators, 7 secondary indicators, and 28 tertiary indicators were determined. Conclusion The evaluation system for the centralized management mode of medical equipment constructed in this study has certain reliability and applicability, and can provide reference value for the evaluation of management modes.

Objective To predict the stage I-II necrosis of the femoral head by a radiomics nomogram based on digital radiography (DR) of the hip joint to expand the application of routine DR in the evaluation of the stage I-II femoral head necrosis. Methods A total of 61 patients with Osteonecrosis of the femoral head (ONFH) and 24 normal healthy subjects were selected as the observation subjects. All the patients in this study received DR and MRI scans of the hip joint. We extracted 1409 radiomics features from all region of interests(ROIs) of the DR images. The minimum absolute contraction and selection operator (LASSO) regression method was used to select the features, and then two machine learning classification models Multilayer Perceptron (MLP) and Support Vector Machine (SVM) were constructed to detect femoral head necrosis. Incorporating radiomics signature(Rad-score) and independent demographics, the radiomics nomogram was established by logistic regression analysis. Receiver operating characteristics curve with area under the curve (AUC), sensitivity, specificity, and sensitivity were used to evaluate diagnostic performance. Results In the validation set, the AUC of the MLP and SVM radiomics models was 0.980 and 0.954, respectively, and the AUC of the radiomics nomogram was 0.981. Conclusion Machine learning based on X-ray radiomics features can assist in screening high-risk individuals with stages I-II femoral head necrosis, ultimately confirmed by MRI images.

Objective By analyzing the current situation and problems of medical device adverse event monitoring and post-market safety evaluation conducted by the first batch of national medical device adverse event monitoring sentinel hospitals (referred as national sentinel hospital) in China. Corresponding suggestions are proposed to provide reference for the construction of management standards for national monitoring sentinel in China. Methods A questionnaire and field survey were conducted among the 105 national sentinel hospitals, and the data analyzed using SPSS 22.0. Results A total of the 92 sentinel hospitals (87.6% of the sample) provided valid responses. The investigation results showed that most sentinel hospitals have established a basic medical device adverse event monitoring system, and conduct monitoring activities in an orderly manner. However, further efforts are needed in risk management and post-market safety evaluation of medical devices. Conclusion Sentinel hospitals should enhance risk management and post-market safety evaluation efforts for medical devices. It is suggested that regulatory authorities should promptly issue relevant guidelines and provide technical training.

Objective This study explores the effectiveness of Supply Processing Distribution (SPD) projects in the acceptance of medical consumables, in order to verify their important role in the refined management of medical consumables. Methods Based on the refined management of SPD project in the acceptance of medical consumables, and summarizes the problems of heavy information input, difficult license supervision, and unclear warehouse location in the original acceptance process of medical consumables. After the implementation of the SPD project, the improvement and optimization results of the acceptance process and information system, such as information scanning and input, electronic license supervision, and location label management, were compared. Statistical methods were used to compare the work effectiveness before and after refined management. Results The results showed that there were significant differences in the acceptance time, abnormal certificate management time, acceptance feedback time, storage and placement time, secondary warehouse acceptance and inventory time, and errors in the acceptance data of the central and secondary warehouses for non direct delivery consumables at different time samples. All indicators showed significance (p<0.01). The acceptance time and errors of different data samples for direct delivery medical consumables (cold chain products, orthopedic implant products) at different time samples showed significance (p<0.05), indicating statistical significance. Conclusion The effectiveness of refined management of medical consumables in the acceptance process has been verified through t-test data, which has improved the efficiency of medical consumables acceptance management in medical institutions and provided scientific and effective suggestions for other medical institutions to achieve refined management of medical consumables.

Objective To address the challenges posed by traditional medical equipment management models in meeting the needs of smart hospitals, especially in terms of equipment usage management and performance evaluation, this study aims to construct a medical equipment usage assessment and management system based Internet of Things(IoT). Methods This study employed IoT technology, integrating the Digital Imaging and Communications in Medicine (DICOM) protocol for medical imaging, Optical Character Recognition (OCR) + Artificial Intelligence (AI) recognition for text extraction from device interfaces, and current bypass detection for monitoring device operational status. These technologies enabled the IoT connectivity of diverse medical equipment across both campuses of the Affiliated Hospital of Jiangnan University. The research also included a comparative analysis of equipment performance before and after the system's implementation using five-point Likert scale and paired t-tests to quantify improvements in equipment performance metrics. Results After the system was launched, a longitudinal analysis was conducted on the performance of 28 large imaging devices. Compared with the same period in 2023, the equipment utilization rate increased by 4.55% (P=0.030), the cost yield rate improved by 15.89% (P=0.002), the fault response time was shortened by 16.88 minutes (P<0.001), and service satisfaction increased by 20.3 points (P<0.001) in 2024. Conclusion The IoT based medical equipment usage assessment and management system effectively solved issues such as data silos, uneven management between two campuses, and data security, significantly enhancing the management level and usage efficiency of medical equipment.

Objective To investigate the clinical value of a dual-flow injection technique combined with low tube voltage scanning based on deep learning algorithm in aortic CTA imaging. Methods A total of 120 patients were divided into 3 groups: A(n=40; tube voltage: 100 kV; conventional injection), B(n=40; tube voltage: 100 kV; dual-flow injection)and C(n=40; tube voltage: 60 kV; dual-flow injection). Group C was divided into two subgroups, C1 was reconstructed by traditional iterative algorithm(Clearview)and C2 was reconstructed by deep learning algorithm(AIClearInfinity), while groups A and B’s algorithm were the same as C1. Objective evaluation(CT, SD, SNR, CNR), subjective scoring, contrast medium dose and patients effective dose (ED) were compared and analyzed among groups. Results The CT values at different aortic measurement levels in group C were higher than those in groups A and B (both P<0.05). There was no statistical significance in SNR and CNR at most levels between groups A and C2 (P>0.05), and the SNR and CNR values in group A were higher than those in groups B and C1 (both P < 0.05). The subjective scores of the image quality in groups A and C2 were higher than those in groups B and C1 (all P<0.05). The contrast medium dose in groups B and C had a 45% reduction compared to group A. Group C’s ED was lower than groups A and B’s (both P<0.05). Conclusion In aortic CTA, a dual-flow injection technique combined with low-tube voltage scanning based on the AIClearInfinity significantly reduces contrast medium dose and radiation dose without compromising image quality.

In order to further optimize and improve the full life cycle management mode of medical consumables, the study used literature research and Delphi method, combined with the application of stakeholders in the medical and health field, to determine the four stakeholders of government agencies, medical institutions, enterprises, and patients, sorted out the current research status and interest demands of the four stakeholders in the full life cycle management of medical consumables, drew a stakeholder influence matrix, further analyzed the conflict points of interests of the four stakeholders, proposed suggestions for the full life cycle management of medical consumables from the perspective of stakeholders based on the conflict of interest demands, further improved the level of medical consumables management, and provided a basis for the formulation and implementation of relevant policies.

Objective To ensure the stable battery status of medical equipment in the deployment center, reduce safety risks, and alleviate the burden on management personnel, a monitoring and control system for the medical equipment deployment center based on electrical energy management is designed. Methods The system uses an electrical energy data acquisition and control module to monitor the charging status of the equipment, and achieves data transmission through Wi-Fi. A TCP server, MySQL database, and Django web platform are built on the PC end to provide real-time monitoring and remote control for management personnel. Results After verification, the system can retrieve the charging status of 8 connected devices after logging in to the web page, and the power on/off operation for the specified device can be completed within 3 seconds after issuing the control command. Compared with the control group, the experimental group showed significant improvements in the frequency of management personnel visiting the allocation center, the time for charging and power off, and the number of devices with battery level ≥50% during inspection. Independent t-test results showed statistically significant differences (t=20.07, 7.28, -0.71, P<0.05). Conclusion This system effectively ensures the stable and safe battery status of the equipment, significantly reduces the workload of clinical engineering personnel, and enhances the emergency response capability of the deployment center. Through the intelligent management of medical equipment power, it provides strong support for the stable operation of hospital departments.

Objective To explore the development and application of a thoracic surgery visualization medical service platform aimed at improving doctor-patient communication and enhancing personalized treatment efficiency. We aim to develop a visualization-based medical service platform to optimize communication between cardiothoracic surgeons and patients,s thereby enhancing the efficiency of personalized treatment. Methods The platform was developed using the Swin-Unet model for multi-organ segmentation of thoracic CT images, combined with the VTK algorithm to generate 3D visualizations. Patient outcomes were compared between an experimental group (using the platform) and a control group (not using the platform), focusing on their understanding of treatment plans, knowledge acquisition rates, consultation duration, and satisfaction levels. The study utilizes the Swin-Unet deep learning model for organ segmentation of chest CT images and integrates the VTK ray-casting algorithm to generate 3D visualization images. An immersive interactive feature was also developed to improve doctor-patient communication. Results Patients in the experimental group showed significantly higher treatment plan understanding scores (4.6±0.5 vs. 3.4±0.7, P < 0.001), knowledge acquisition rates (72.6±6.2% vs. 38.7±8.5%, P < 0.001), and satisfaction levels (4.7±0.4 vs. 3.8±0.6, P < 0.001) compared to the control group. Consultation time was also significantly reduced (12.4±2.1 minutes vs. 18.2±3.5 minutes, P < 0.001).The platform was successfully applied to the personalized 3D reconstruction of 1,382 chest CT images. Patients in the experimental group showed significantly higher scores in understanding treatment plans, knowledge retention, and satisfaction compared to the control group, with a substantial reduction in communication time (P<0.001). Conclusion The platform effectively improved diagnostic efficiency and doctor-patient communication, providing a new tool for personalized precision medicine. Future development could integrate virtual reality technology to further expand its application in clinical settings.This study demonstrated the potential of artificial intelligence and visualization technologies in enhancing personalized treatment in thoracic surgery. The platform provides valuable insights and technical support for building future multidisciplinary collaboration frameworks in precision medicine.

Addressing the issue of the high concealment of fault features in medical imaging equipment, where a single SVM method struggles to effectively extract deep-level fault features, leading to a low Kappa coefficient in fault classification, a fault classification method for CT medical imaging equipment based on Particle Swarm Optimization (PSO) SVM is proposed. Multiple types of sensors were utilized to monitor the operational status signals of CT medical imaging equipment in real-time. The collected signals were processed through wavelet transform to remove noise and extract valid features. The processed signals were then input into a Deep Belief Network (DBN). The DBN, by stacking Restricted Boltzmann Machines (RBMs) layer by layer and undergoing two stages of unsupervised training and supervised parameter tuning, precisely captured and learned the deep-level fault features of CT medical imaging equipment. The extracted fault features were fed into a multi-classifier based on PSO-SVM, and CT medical imaging equipment fault classification was achieved through the trained model. Experimental results demonstrate that, with the DBN configured with 4 layers and 80, 150, and 80 neurons in the input, hidden, and output layers respectively, the proposed method achieved an F1 score of 0.925, a Kappa coefficient of 0.0895, and a Hamming distance below 0.053. These validation data indicate that the proposed method can accurately classify faults in CT medical imaging equipment, providing powerful technical support for fault diagnosis and maintenance of medical imaging equipment.

Objective To investigate the value of single-energy imaging, energy spectrum curve and atomic number of energy spectrum CT in the diagnosis of negative gallbladder stones. Methods The imaging data of 76 patients with suspected negative gallstones were retrospectively collected, and all patients underwent cholecystectomy, with the surgical results as the reference standard.The differences of single energy level 40 keV images, energy spectrum curve characteristics and effective atomic number between negative gallstones and bile were compared.The diagnostic efficacy of single energy level 40 keV image, energy spectrum curve and effective atomic number for negative stones and bile in gallbladder was analyzed by ROC curve.Results The mean atomic number at 40 keV was 6.12±0.37 for negative gallbladder stones and 7.45±0.41 for bile, and the difference was statistically significant (P<0.05). The CT value of negative gallbladder stones was -40.12±12.67Hu and the CT value of bile was 5.78±14.89Hu, and the difference was statistically significant (P<0.05). The slopes of the energy spectrum curves were -1.58±0.39 for negative gallbladder stones and 0.27±0.12 for bile, and the differences were statistically significant (P<0.05).The ROC curves showed areas under the curve of 0.94, 0.951, and 0.97 for the single-energy-level 40 keV image, the energy-spectral curve, and the atomic number map, with sensitivities and specificities of 88.89% and 84.21%, 92.31% and 91.12%, and 100% and 96.11%, respectively.Conclusion Energy-spectrum CT multiparametric imaging helps to detect gallbladder stones at an early stage, guides clinical therapeutic decisions, reduces the risk of misdiagnosis and omission, and has a positive impact on the treatment and prognosis of patients.

Objective To propose a surgical instrument identification method based on the improved YOLOv8n model, aiming to reduce the errors that are prone to occur in traditional manual counting methods and further protect the safety of patients. Methods Firstly, the SE attention mechanism is introduced in the backbone of the network to enhance the ability of the model to utilize the feature information. Secondly, the model incorporates the BiFPN pyramid structure to more effectively integrate multi-scale features. Finally, WIoU loss function is used to optimize the original network loss function, so as to improve the accuracy of the model. Results The experimental results showed that the performance of the improved YOLOv8n model was significantly better than the original YOLOv8n model in the task of surgical instrument recognition, with mAP50/% and MAP50-95 /% improved by 6.8% and 8%, respectively. Conclusion The improved model is also superior to common algorithms such as YOLOv4 in terms of accuracy, effectively reducing the phenomenon of missing instrument detection, and significantly improving the reliability and accuracy of surgical instrument identification.

Objective Visual analysis tools were used to explore and summarize the research status, hot spots and trends of SPD management mode in medical consumables. Methods With the core database of China National Knowledge Network (CNKI) as the literature data source and the literature published so far in the field of SPD management mode as the main analysis object, the CiteSpace 6.3.R1 software was used to analyze the literature publication trend, authors, research institutions and keywords in this field, and the related visual maps were drawn. Results The relevant academic achievements were first published in 2016, and the number of papers has been on the rise in recent years. Most of the research institutions are concentrated in the southern region, while the research institutions in the northern region are relatively few. The highest frequency of occurrence is "medical supplies" (89 times), "fine management" (45 times),"medical supplies management" (19 times), and the central value of such keywords is also high. The research on the SPD distribution model for medical consumables in China began in 2016, which is relatively late. From 016 to 2019 was the initial stage of the SPD distribution model, mainly focusing on the changes in hospital logistics management, and replacing the methods of entry and exit from the warehouse, acceptance, and requisition through the SPD model. In recent years, hospitals have begun to pay attention to the refined of medical consumables, including automatic reordering, real-time tracking, and rapid response, and gradually extended from management to management plus supervision. For medical institutions with campuses, the SPD model helps to reduce duplicate construction. Conclusion The SPD model has great potential in optimizing medical material management and improving hospital operational efficiency while ensuring data security.

Transcranial magnetic stimulator is a noninvasive neuromodulation device, which has been widely used in many fields such as psychiatry and neurology, but prolonged operation is prone to lead to its failure. This paper describes three cases of Magneuro 100 transcranial magnetic stimulator failure. The first case is overheating alarm, first analyze the heat dissipation mechanism, determine the heat dissipation fan is damaged, replace the problem solved. The second case is capacitor voltage abnormality, by analyzing the charging principle of the high-voltage storage capacitor, disassembling the machine and testing the key electronic components, and found that the insulated gate bipolar transistor was damaged, and the problem was solved after replacement. The third case is the failure of the motor evoked potential monitoring module, EMG can not be displayed, after the signal acquisition and transmission path inspection, found that the synchronization interface and the circuit board of the weld joints loose, re-welded after the problem is solved. This paper can help equipment engineers to master the structural principles and troubleshooting methods of similar equipment to protect the continuity and safety of hospital work.

Objective To solve the problem of low maintenance quality and efficiency in traditional modes, provide an online digital platform for operation training and troubleshooting of various medical equipment, and improve the quality and efficiency of medical equipment maintenance. Methods Adopting front-end and back-end separation technology, building a technical support auxiliary system for medical equipment training, troubleshooting, and quality control based on Python-Django framework and VUE framework. The system supports mobile phone usage and can be used as a medical device teaching system. It also has evaluation and feedback functions, which can dynamically optimize technical guidance solutions. Results After the system trial operation, the average daily number of equipment repairs increased from 7.53 to 9.43, and the average daily fault repair rate increased from 78% to 86%, with statistical significance (P<0.05). Conclusion this system has the characteristics of medical equipment composition in our hospital. By digitizing technical support data and providing professional guidance through a digital platform, it can help improve the level of maintenance management. It is called an iteratively updated "maintenance knowledge base".

Objective This study aims to analyze the operating costs of laboratory department by constructing a cost model for laboratory tests, so as to achieve more effective cost control and resource optimization. Methods The operating data of the immunology group and biochemistry group of the laboratory department were collected, including the income and reagent cost of each kind of test, and the total income, theoretical cost rate, actual cost rate, non-profit cost rate, and the deviation between the theoretical and actual cost rates were calculated. The cost structure of two groups of test projects through the correlation between deviation and non-measured cost rate was analyzed. Results The actual and theoretical cost rates of the biochemical group were significantly lower than 40%, and the actual cost rate of the immune group was higher than 40%, but there was no significant difference between the theoretical cost rate and 40%. The high cost was mainly caused by the increase of quality control and calibration costs in the biochemical group. It could be improved by reducing the frequency of reagent batch replacement. In biochemistry group, the high cost is mainly caused by additional depletion of reagents and the imprecision of the theoretical cost rate measurement. The number of theoretical tests of reagents should be carefully evaluated. Conclusion The application of cost accounting model is not only helpful to identify and analyze the cost structure of laboratory tests, but also provides an important reference for further optimizing resource allocation and improving operational efficiency of the laboratory department.

With a history for more than 200 years, blood gas analysis has been widely used in clinical medicine. Currently, the commonly used methods for blood gas analysis are invasive detection technology. Although the technology is extensively employed, there are also many problems, such as infection, bleeding due to the invasive operation and discontinuous mointroing. Therefore, a number of non-invasive blood gas monitoring technologies have gradually emerged to make up for the shortcomings. In this paper, we reviewed the development history of blood gas analysis technology in detail, and we also discussed the development of non-invasive blood gas analysis technology. With a special attention to the application of optical Cavity Ring-Down Spectroscopy (CRDS) in non-invasive blood gas analysis and monitoring technology, we believe this is a direction for the future development of non-invasive blood gas technology.

Assisted medical diagnosis has become an important research direction in deep learning. Based on multi-layer neural networks, non-linear mapping between input and specific output can be achieved by extracting shallow features of combined data, especially in capturing lesion area features and revealing details that are difficult to observe with the naked eye. This article mainly focuses on the application of deep learning networks in the precise automatic classification of gastric cancer, the relationship between deep features and genes, and prognosis evaluation, with a focus on sorting out the basic ideas and improvement points of the model; Providing new perspectives and methods for computer-aided diagnosis has brought different entry points for accurate diagnosis and personalized treatment of patients.

Objective To improve the in-hospital procurement process of orthopaedic consumables, to better complete the task of quantity procurement, to achieve fine management of orthopaedic consumables, and to promote the high-quality development of public hospitals. Methods Based on SPD system, the information island of each department was opened up, process reengineering was promoted, and the integrated whole process management of orthopedics was formed, and the task of contract group was completed with information tools. Results The standardization of orthopedic implant management was significantly improved, the satisfaction of patients was significantly improved, the accuracy of disinfection kits in disinfection and supply room was significantly improved, and the statistical time of set tape volume was significantly shortened, with statistical significance (P < 0.05). Conclusion The application of SPD system can improve the management efficiency of hospital, reduce the operating cost of hospital, ensure the smooth implementation of volume procurement, and provide data support for hospital managers to make scientific decisions.

Objective To reveals the research trends and vital topics in the field of medical engineering by conducting a bibliometric analysis on the historical literature data. Methods The PubMed database and the Medpulse bibliometric analysis platform were used to retrieve medical engineering research literature published from 2014 to 2024. The retrieved literature was subjected to data mining using the R package 'bibliometrix' (version 4.2.3). Results A total of 3,279 medical-engineering-related papers were retrieved, cited 2,589 times in total, with the peak publication year being 2023, which saw 442 articles published. The journal HELIYON featured the most articles on the subject, with a total of 138 publications. The two most prolific authors are from the United States and China. The top three research institutions by publication volume are all located in the United States. Both the United States and China have actively participated in international collaborations in the field of medical engineering. Key research themes over the past decade include “Biomedical Engineering” “Tissue Engineering” and “Drug Delivery”. The topics of high impact factor articles are relatively concentrated, while the topics of highly cited articles are diverse. Conclusion This study provides a comprehensive analysis of the hot topics and development trends in the field of medical engineering over the past decade, offering a solid foundation for professionals to understand their positioning and select research topics.

Objective To detect the continuity of three modules in the software by using TQA tool of Tomotherapy, and to observe and analyze the stability of dynamic Jaw. MethodsThe data required for the three TQA modules were collected using the Radixact detector and ionization chamber, field width constancy (FWC) and Maximum Jaw Encoder Error (MJEE) were used to observe the field width constancy (J01 and J07) and dynamic Jaw constancy in Daily QA module ;In the Field Width-Dynamic JAWS module, the Gamma Index Maximum (GIM) and the Field Width Percent Difference (FWPD) are used to measure the stability of four fields (J42, J20, J14, J07) in symmetric and asymmetric conditions ;The Time Skew (TS) and Relative Jaw fluence output factor (RJFOF) were calculated in Jaw Sweep-Dynamic Jaws module to observe the field width stability, Dynamic lead gate Time accuracy, Dynamic Jaw scanning stability and Dynamic lead gate speed accuracy. Results The average FWC of J01 in the Daily QA module is 98.88% ± 0.67%, which is within the normal range of 95-105 and shows a relatively stable state. The average FWC of J07 is 99.75% ± 0.17%, which is within the normal range of 98-102. The fluctuation range of MJEE of the front and back jaw doors over time is -0.3 to 0.2, with an average value of micrometers, indicating that the motion error of the front and back jaw is very small. The ten states of four different jaw widths in the Field Width Dynamic Jaws module vary over time. The GIM of FW is within the normal tolerance range of 0-1, and the FWPD is also within the tolerance range of -1 to 1 over time. The fluctuation of jaw width over time is relatively small. The jaw reaction speed in the Jaw Sweep Dynamic Jaws module remains stable over time, with an average of -0.006 seconds ± 0.003 seconds. The average values of RJFOF J20, RJFOF J14, and RJFOF J07 are 0.999757 ± 0.000557, 0.999844 ± 0.000662, and 0.998641 ± 0.001008, respectively, indicating that there is almost no change in the output factor values of the injection volume. The corresponding field widths in the three modules are within the tolerance range, with precision errors in the micrometer level. The jaw response time is extremely short, and the RJFOF value is close to the ideal value of 1. Conclusion Through the analysis of one-year data of three modules, the dynamic Jaw is stable with time.

Objective Diffusion correlation tomography (DCT) can detect blood flow changes induced by early tissue disease. As the tomographic imaging device, large-area contact DCT equipment requires a large-scale source detector (S-D) array to complete data acquisition, resulting in high equipment costs. Methods To address this issue, this article proposes using an optical switch array to time-division multiplex one S-D group into eight groups, covering a total of 8×8 cm2 of target tissue. The S-D cross distribution of long and short distances in each group forms a concentric circular structure. By utilizing its symmetry and equivalence, the balance of photons obtained by each voxel is ensured, and the detection area and depth is expanded. Results Based on this low-cost DCT device, the reconstruction position of the cross-shape phantom anomaly is accurate, and the morphology is almost complete. The contrast of the second reconstructed slice (0.5-1 cm depth) is 0.75. The contrast of the tubular anomaly is well proportional to the flow rate, and the reconstruction position is accurate, the shape is complete. In the cuff compression clinical test, the blood flow index in the relaxed state is about 10 times that of the compression state. The 3D reconstructed image and its standard deviation also indicates that the blood flow in the relaxed state is more abundant. Conclusion By the time-division multiplexing function of the optical switch, the cost of this contact DCT device with large detection area has been effectively reduced by 60%, and the detection depth is almost 1.5 cm. The experimental results also indicate that this contact DCT device has powerful detection ability for different morphological abnormalities, and can be used as a new detection method to detect diseases related to abnormal blood flow perfusion.

Objective This study visualizes the past decade of research on resting-state functional magnetic resonance imaging (rs-fMRI), highlighting key trends and emerging topics. Methods English rs-fMRI literature from 2014 to 2023 were retrieved from the Web of Science Core Collection. VOSviewer and CiteSpace software were used to analyze and map the data. Results A total of 10,581 articles were analyzed, showing a steady increase in publications. China has the highest number of publications (4836 papers), while the United States has the highest number of citations (111,945 citations). Capital Medical University and Gong Qiyong were the top institution and author, respectively. 《Frontiers in Neuroscience》 published the most articles. Frequent keywords included functional connectivity, default mode network, and brain network, with recent trends focusing on machine learning and cognitive function. Conclusions In recent years, Interest in rs-fMRI research is growing. Analytical methods have transitioned from traditional techniques like functional connectivity to innovative approaches such as degree centrality and dynamic functional connectivity. The focus has shifted from neuroimaging features in brain diseases to the integration of imaging and artificial intelligence for accurate diagnosis and classification. Future studies will likely emphasize optimizing data analysis methods and expanding the applications of machine learning technologies.

Due to the complex operating environment of medical equipment, there are many interference factors in the operating status data, which affect the accuracy of equipment status monitoring. Therefore, a monitoring method for abnormal operation status of medical equipment based on multi-source data analysis is proposed. Using sensors to collect operational data from multiple sources of medical equipment, first calculate the soft threshold of operational status and determine the fluctuation range of operational status data. Next, use the correlation matrix to fuse the operational status data of multi-source medical equipment. Based on the results of data fusion, extract time-domain and frequency-domain features of the data, establish data feature vectors, and achieve feature mining of the data state. Then, by calculating the characteristic entropy value and operating state value of the running state data, it is determined whether the device is in an abnormal state. Finally, calculate the ratio of abnormal states of medical equipment to classify different types of abnormal states. The experimental results show that the average false alarm rate of this method in practical applications is 3.25%, demonstrating high monitoring accuracy; Can effectively improve monitoring accuracy and reduce false alarm rates.