A multi-layer feature attention enhanced network is proposed to further improve the diagnostic accuracy of the severity of diabetic retinopathy.To address the inconsistent expression of global and local features when processing diabetic retinopathy images,a dual-branch parallel model combining ResNet-50 and DeiT-S is employed as the backbone architecture,and a feature fusion module is designed at the end of the network.Concurrently,a multi-scale location awareness enhancement module is developed to extract multi-scale information through dilated convolution with positional attention mechanism for enhancing the feature representation of lesions in fundus images,and a local feature enhancement module is constructed to strengthen the model's capability in extracting local information,thus improving model's capability to identify small lesions and minor changes.The experimental results show that the proposed multi-layer feature attention enhanced network achieves an accuracy of 87.61%,exhibiting excellent classification performance.This advancement provides a strong support for further development of diabetic retinopathy detection technology.

Objective:To build a big data analysis model of comprehensive performance of medical equipment,so as to improve the management efficiency and fine management level for medical equipment of hospital.Methods:The big data analysis model of comprehensive performances of medical equipment was developed through building a data integration platform,and integrating multi-business system data of hospital,and including quantified social benefit data and relative data of the process of procurement and maintenance,which established index system of performance analysis with multi-dimensions and multi-layers.It realized the presentation of evaluation results of performance of medical equipment in the form of visual data reports.Results:The big data analysis model for the comprehensive performance of medical equipment has set a system with more than 10 indicators,which included social-benefit indicators,evaluation indicators of comprehensive performance,income indicators,and the indicator of recovery period of investment,and visual report with 35 pages.Each layer of the performance of equipment was analyzed from 35 angles,which realized a comprehensive,multi-dimensional and refined evaluation for the performance of medical equipment of hospital.Its'agile,easy-to-use,efficient response and automation features can significantly improve work efficiency of business.The model realized automatic data collection through interfaces and other means,and reduced manual input errors,which data accuracy rate was＞99%.The results of testing the performance data of equipment for three consecutive years by using this model indicated that the consuming time of single calculation of business system of hospital was shortened from＞20 min to 5-8 s.Conclusion:The big data analysis model of comprehensive performance of medical equipment can help hospital to identify,prevent and control risks in advance,and improve the efficiency of internal control,and enhance management efficiency and fine management level for hospital,and provide support for management decision for medical equipment of hospital.

A multi-layer feature attention enhanced network is proposed to further improve the diagnostic accuracy of the severity of diabetic retinopathy.To address the inconsistent expression of global and local features when processing diabetic retinopathy images,a dual-branch parallel model combining ResNet-50 and DeiT-S is employed as the backbone architecture,and a feature fusion module is designed at the end of the network.Concurrently,a multi-scale location awareness enhancement module is developed to extract multi-scale information through dilated convolution with positional attention mechanism for enhancing the feature representation of lesions in fundus images,and a local feature enhancement module is constructed to strengthen the model's capability in extracting local information,thus improving model's capability to identify small lesions and minor changes.The experimental results show that the proposed multi-layer feature attention enhanced network achieves an accuracy of 87.61%,exhibiting excellent classification performance.This advancement provides a strong support for further development of diabetic retinopathy detection technology.

Objective:To build a big data analysis model of comprehensive performance of medical equipment,so as to improve the management efficiency and fine management level for medical equipment of hospital.Methods:The big data analysis model of comprehensive performances of medical equipment was developed through building a data integration platform,and integrating multi-business system data of hospital,and including quantified social benefit data and relative data of the process of procurement and maintenance,which established index system of performance analysis with multi-dimensions and multi-layers.It realized the presentation of evaluation results of performance of medical equipment in the form of visual data reports.Results:The big data analysis model for the comprehensive performance of medical equipment has set a system with more than 10 indicators,which included social-benefit indicators,evaluation indicators of comprehensive performance,income indicators,and the indicator of recovery period of investment,and visual report with 35 pages.Each layer of the performance of equipment was analyzed from 35 angles,which realized a comprehensive,multi-dimensional and refined evaluation for the performance of medical equipment of hospital.Its'agile,easy-to-use,efficient response and automation features can significantly improve work efficiency of business.The model realized automatic data collection through interfaces and other means,and reduced manual input errors,which data accuracy rate was＞99%.The results of testing the performance data of equipment for three consecutive years by using this model indicated that the consuming time of single calculation of business system of hospital was shortened from＞20 min to 5-8 s.Conclusion:The big data analysis model of comprehensive performance of medical equipment can help hospital to identify,prevent and control risks in advance,and improve the efficiency of internal control,and enhance management efficiency and fine management level for hospital,and provide support for management decision for medical equipment of hospital.

Fibromyalgia syndrome （FMS） is a refractory， chronic non-articular rheumatic disease characterized by widespread pain throughout the body， for which there are no satisfactory therapeutic drugs or options. There are rich Chinese medical therapies， and some non-drug therapies， such as acupuncture， Tai Chi， and Ba-Duan-Jin， have shown satisfactory efficacy and safety and definite advantages of simultaneously adjusting mind and body. FMS is taken as a disease responding specifically to traditional Chinese medicine （TCM） by the National Administration of Traditional Chinese Medicine in 2018. In order to clarify the research progress in FMS and the clinical advantages of TCM/integrated Chinese and Western medicine， the China Academy of Chinese Medicine organized a seminar for nearly 20 experts in Chinese and Western medicine， including rheumatology， psychology， acupuncture and moxibustion， and encephalopathy， with the topic of difficulties in clinical diagnosis and treatment of FMS and advantages of TCM and Western medicine. The recommendations were reached on the difficulties in early diagnosis and solutions of FMS， mitigation of common non-specific symptoms， preferential analgesic therapy， TCM pathogenesis and treatment advantages， and direction of treatment with integrated Chinese and Western medicine. FMS is currently facing the triple dilemma of low early correct diagnosis， poor patient participation， and unsatisfactory benefit from pure Western medicine treatment. To solve the above problems， this paper suggests that rheumatologists should serve as the main diagnostic force of this disease， and they should improve patient participation in treatment decision-making， implement exercise therapy， and fully utilize the holistic and multidimensional features of TCM， which is effective in alleviating pain， improving mood， and decreasing adverse events. In addition， it is suggested that FMS treatment should rely on both TCM and Western medicine and adopt multidisciplinary joint treatment， which is expected to improve the standard of diagnosis and treatment of FMS in China.

Concurrent chemoradiotherapy (CCRT) refers to the simultaneous treatment of chemotherapy and radiotherapy, and the effect of radiotherapy is enhanced with low-dose chemotherapy, which can reduce tumor recurrence and metastasis and improve clinical prognosis of patients. At present, the main factors for the increase of radiosensitivity of concurrent chemotherapy is that concurrent chemotherapy prevents the repair of tumor cells, and chemotherapy and radiotherapy act on different cell cycles and have synergistic effects. However, even for patients with locally advanced cervical cancer (LACC) who have undergone CCRT, the 5-year survival rate is only 60%, which is still not ideal. In order to improve the efficacy, researchers have conducted a series of exploratory studies, which consist of the combination of targeted drugs and immunodrugs, and neoadjuvant regimens before CCRT, etc. Although targeted or immunologic drugs are effective treatment of LACC, in view of the lack of large-scale evidence-based medical evidence, multi-center prospective and randomized phase III clinical trials and high-level articles are needed to improve the level of evidence-based medicine. This consensus summarizes several key evidence-based medical studies published recently, especially the clinical research progress in targeted and immunological therapies, providing reference for domestic peers.

Objective:To explore the application and potential optimization of the collaborative and competitive learning model in the Health Education course. Methods:Undergraduate medical students participating in Health Education course practice tasks were selected to conduct discussions and reach consensus according to research objectives based on the e fast anonymous consensus-forming tool (eFAST). The meeting records were analyzed for theme identification using the keyword classification method. Results:Nine medical students participated in eFAST discussions. The students considered the following five aspects as the most important for undertaking Health Education course practice tasks using the collaborative and competitive learning model: timely communication, problem evaluation, report content enrichment, reasonable task allocation within groups, and task topic selection by group members together. They also proposed suggestions on improvement of the assessment method, including teacher involvement in scoring, intra-group scoring based on inter-group scoring, all students participating in inter-group scoring, and using mobile applications for scoring and summarization. Conclusions:The collaborative and competitive learning model can be used in the teaching of Health Education, but further optimization is needed in course task design, implementation, reporting, and assessment.

Objective　To understand the impact of early and timely treatment and initial antiviral treatment regimen on mortality and attrition of antiretroviral therapy. Methods A retrospective cohort study was conducted using download data on antiretroviral therapy for HIV-infected patients in Liuzhou City, Guangxi Province, from the database of the Basic Information System for AIDS Control and Prevention (BISAC) from 2010 to 2020. The Cox proportional risk regression model was used to analyze the influencing factors of mortality and attrition. Results A total of 15 713 infected patients were included, including 53.4% aged 18-<50 years, 69.4% male, 61.0% farmer, 75.1% CD4 count <350 cells /μL before initial antiviral treatment, the overall mortality rate was 4.30/100 person-years, and the overall attrition was 2.42/100 person-years. The results of Cox regression analysis showed that the influencing factors of mortality were pretreatment CD4 counts of 350-<500 cells/μL(AHR=0.72, 95%CI: 0.63-0.81) and ≥500 cells/μL (AHR= 0.64, 95%CI: 0.55-0.76); duration from diagnosis to initial antiviral treatment 91-180 days (AHR=1.25, 95%CI: 1.08-1.45), 181-365 days (AHR=1.26, 95%CI: 1.08-1.47), and ≥365 days (AHR=1.26, 95%CI: 1.11-1.44); initial antiviral treatment regimens of D4T+3TC+EFV/NVP (AHR=1.47, 95%CI: 1.32-1.63) and AZT/D4T/TDF+3TC+LPV/r (AHR=1.73, 95%CI: 1.50-1.99). Factors affecting attrition were pretreatment CD4 counts of 350-499 cells/μL (AHR=1.32, 95%CI: 1.16-1.50) and ≥500 cells/μL (AHR=1.28, 95%CI: 1.10-1.50); interval from HIV positivity confirmation to initial dosing ≥365 days (AHR=1.21, 95%CI: 1.04-1.40), initial antiviral treatment regimens of TDF+3TC+NVP (AHR=1.32, 95%CI: 1.13-1.55), AZT+3TC+EFV/NVP (AHR=1.43, 95%CI: 1.26-1.62) and AZT/D4T/TDF+3TC+LPV/r (AHR=1.33, 95CI%: 1.06-1.67). Conclusions　Early and timely treatment and the initial antiviral treatment regimen of TDF+3TC+EFV have good efficacy, but attention should be paid to the high risk of attrition of HIV-infected people with high CD4 count before treatment.

Objective:To understand the positive rate of 2019 novel coronavirus (2019-nCoV) specific IgG antibody induced by Coronavirus Disease 2019 (COVID-19) inactivated vaccine in healthy population in Xizang Autonomous Region, and evaluate the immune effect of the vaccine.Methods:Serum samples were collected from COIVD-19 vaccine immunized health population without history of 2019-nCoV infection from six prefecture-level cities in Xizang Autonomous Region. The IgG antibody against 2019-nCoV were tested by chemiluminescence method. Then, the positive rate of IgG antibody was analyzed for different immunization histories and age groups.Results:A total of 22 255 participants were enrolled in this survey. After full-access (two doses of vaccine) and booster immunization, the overall positive rate of specific IgG antibody against 2019-nCoV was 96.38%. The positive rate of IgG antibody in the booster immunized population was 97.12%, which was much higher than the 88.38% in the full-access immunization population, the difference is statistically significant ( χ2=381.11, P<0.001). There was a significant differences in the positive rates of specific IgG antibodies in different age groups ( χ2=138.28, P<0.001). Especially in the younger age groups, including less than 10 years old and the 11-20 years age group, the positive rate of specific IgG antibody were 93.44% and 89.03% respectively, which were lower than those in other age groups. Except for Naqu city and the age group ≤ 10 years old, the differences in antibody positivity rates were statistically significant between different age groups in the same region and between the different regions in the same age group for the two populations with different immunization histories ( P<0.05). Conclusions:The COVID-19 inactivated vaccine showed a good immune effect in the healthy population in Xizang Autonomous Region, and the booster immunization helps to increase the positive rate of specific IgG antibody in healthy population.

The development of information technology in the medical industry is accelerating the construction of 3-level rehabilitation networks centering on community-based rehabilitation, which is based on the Internet and in combination with the Internet of things(IOT). A municipal rehabilitation hospital cooperated with the health centers of sub-district communities in the city to build a remote rehabilitation network based on IOT technology, for rehabilitation training of community stroke patients. The two sides worked out a cooperation scheme on project management, designed and built a remote community rehabilitation management model. Under the remote monitoring and equipment data IOT, the medical team of the municipal rehabilitation hospital could formulate the therapeutic plan according to the immediately obtained relevant functional evaluation data, and the medical team of the community health service center was responsible for the implementation of the plan. Fifty-three patients in the experimental group received the training of remote lower limb intelligent feedback system based on IOT together with routine rehabilitation training. After 8 weeks of treatment, the functional evaluation data of the two groups were improved in varying degrees( P< 0.05), but the improvement of the experimental group was significantly better than that of the control group( P < 0.05). Remote rehabilitation under the construction path of remote rehabilitation network based on IOT technology could significantly promote the improvement of post-functional rehabilitation of community stroke patients. This study can provide reference for the construction of three-level remote rehabilitation networks based on IOT and the implementation of home-based remote rehabilitation therapy in the future.