Objective: This study aims to overcome challenges in lumbar spine imaging, particularly lumbar spinal stenosis, by developing an automated segmentation model using advanced techniques. Traditional manual measurement and lesion detection methods are limited by subjectivity and inefficiency. The objective is to create an accurate and automated segmentation model that identifies anatomical structures in lumbar spine magnetic resonance imaging scans. Methods: Leveraging a dataset of 539 lumbar spinal stenosis patients, the study utilizes the residual U-Net for semantic segmentation in sagittal and axial lumbar spine magnetic resonance images. The model, trained to recognize specific tissue categories, employs a geometry algorithm for anatomical structure quantification. Validation metrics, like Intersection over Union (IOU) and Dice coefficients, validate the residual U-Net’s segmentation accuracy. A novel rotation matrix approach is introduced for detecting bulging discs, assessing dural sac compression, and measuring yellow ligament thickness. Results: The residual U-Net achieves high precision in segmenting lumbar spine structures, with mean IOU values ranging from 0.82 to 0.93 across various tissue categories and views. The automated quantification system provides measurements for intervertebral disc dimensions, dural sac diameter, yellow ligament thickness, and disc hydration. Consistency between training and testing datasets assures the robustness of automated measurements. Conclusion Automated lumbar spine segmentation with residual U-Net and deep learning exhibits high precision in identifying anatomical structures, facilitating efficient quantification in lumbar spinal stenosis cases. The introduction of a rotation matrix enhances lesion detection, promising improved diagnostic accuracy, and supporting treatment decisions for lumbar spinal stenosis patients.

Objective: This study aims to overcome challenges in lumbar spine imaging, particularly lumbar spinal stenosis, by developing an automated segmentation model using advanced techniques. Traditional manual measurement and lesion detection methods are limited by subjectivity and inefficiency. The objective is to create an accurate and automated segmentation model that identifies anatomical structures in lumbar spine magnetic resonance imaging scans. Methods: Leveraging a dataset of 539 lumbar spinal stenosis patients, the study utilizes the residual U-Net for semantic segmentation in sagittal and axial lumbar spine magnetic resonance images. The model, trained to recognize specific tissue categories, employs a geometry algorithm for anatomical structure quantification. Validation metrics, like Intersection over Union (IOU) and Dice coefficients, validate the residual U-Net’s segmentation accuracy. A novel rotation matrix approach is introduced for detecting bulging discs, assessing dural sac compression, and measuring yellow ligament thickness. Results: The residual U-Net achieves high precision in segmenting lumbar spine structures, with mean IOU values ranging from 0.82 to 0.93 across various tissue categories and views. The automated quantification system provides measurements for intervertebral disc dimensions, dural sac diameter, yellow ligament thickness, and disc hydration. Consistency between training and testing datasets assures the robustness of automated measurements. Conclusion Automated lumbar spine segmentation with residual U-Net and deep learning exhibits high precision in identifying anatomical structures, facilitating efficient quantification in lumbar spinal stenosis cases. The introduction of a rotation matrix enhances lesion detection, promising improved diagnostic accuracy, and supporting treatment decisions for lumbar spinal stenosis patients.

Objective: This study aims to overcome challenges in lumbar spine imaging, particularly lumbar spinal stenosis, by developing an automated segmentation model using advanced techniques. Traditional manual measurement and lesion detection methods are limited by subjectivity and inefficiency. The objective is to create an accurate and automated segmentation model that identifies anatomical structures in lumbar spine magnetic resonance imaging scans. Methods: Leveraging a dataset of 539 lumbar spinal stenosis patients, the study utilizes the residual U-Net for semantic segmentation in sagittal and axial lumbar spine magnetic resonance images. The model, trained to recognize specific tissue categories, employs a geometry algorithm for anatomical structure quantification. Validation metrics, like Intersection over Union (IOU) and Dice coefficients, validate the residual U-Net’s segmentation accuracy. A novel rotation matrix approach is introduced for detecting bulging discs, assessing dural sac compression, and measuring yellow ligament thickness. Results: The residual U-Net achieves high precision in segmenting lumbar spine structures, with mean IOU values ranging from 0.82 to 0.93 across various tissue categories and views. The automated quantification system provides measurements for intervertebral disc dimensions, dural sac diameter, yellow ligament thickness, and disc hydration. Consistency between training and testing datasets assures the robustness of automated measurements. Conclusion Automated lumbar spine segmentation with residual U-Net and deep learning exhibits high precision in identifying anatomical structures, facilitating efficient quantification in lumbar spinal stenosis cases. The introduction of a rotation matrix enhances lesion detection, promising improved diagnostic accuracy, and supporting treatment decisions for lumbar spinal stenosis patients.

Objective: This study aims to overcome challenges in lumbar spine imaging, particularly lumbar spinal stenosis, by developing an automated segmentation model using advanced techniques. Traditional manual measurement and lesion detection methods are limited by subjectivity and inefficiency. The objective is to create an accurate and automated segmentation model that identifies anatomical structures in lumbar spine magnetic resonance imaging scans. Methods: Leveraging a dataset of 539 lumbar spinal stenosis patients, the study utilizes the residual U-Net for semantic segmentation in sagittal and axial lumbar spine magnetic resonance images. The model, trained to recognize specific tissue categories, employs a geometry algorithm for anatomical structure quantification. Validation metrics, like Intersection over Union (IOU) and Dice coefficients, validate the residual U-Net’s segmentation accuracy. A novel rotation matrix approach is introduced for detecting bulging discs, assessing dural sac compression, and measuring yellow ligament thickness. Results: The residual U-Net achieves high precision in segmenting lumbar spine structures, with mean IOU values ranging from 0.82 to 0.93 across various tissue categories and views. The automated quantification system provides measurements for intervertebral disc dimensions, dural sac diameter, yellow ligament thickness, and disc hydration. Consistency between training and testing datasets assures the robustness of automated measurements. Conclusion Automated lumbar spine segmentation with residual U-Net and deep learning exhibits high precision in identifying anatomical structures, facilitating efficient quantification in lumbar spinal stenosis cases. The introduction of a rotation matrix enhances lesion detection, promising improved diagnostic accuracy, and supporting treatment decisions for lumbar spinal stenosis patients.

Objective: This study aims to overcome challenges in lumbar spine imaging, particularly lumbar spinal stenosis, by developing an automated segmentation model using advanced techniques. Traditional manual measurement and lesion detection methods are limited by subjectivity and inefficiency. The objective is to create an accurate and automated segmentation model that identifies anatomical structures in lumbar spine magnetic resonance imaging scans. Methods: Leveraging a dataset of 539 lumbar spinal stenosis patients, the study utilizes the residual U-Net for semantic segmentation in sagittal and axial lumbar spine magnetic resonance images. The model, trained to recognize specific tissue categories, employs a geometry algorithm for anatomical structure quantification. Validation metrics, like Intersection over Union (IOU) and Dice coefficients, validate the residual U-Net’s segmentation accuracy. A novel rotation matrix approach is introduced for detecting bulging discs, assessing dural sac compression, and measuring yellow ligament thickness. Results: The residual U-Net achieves high precision in segmenting lumbar spine structures, with mean IOU values ranging from 0.82 to 0.93 across various tissue categories and views. The automated quantification system provides measurements for intervertebral disc dimensions, dural sac diameter, yellow ligament thickness, and disc hydration. Consistency between training and testing datasets assures the robustness of automated measurements. Conclusion Automated lumbar spine segmentation with residual U-Net and deep learning exhibits high precision in identifying anatomical structures, facilitating efficient quantification in lumbar spinal stenosis cases. The introduction of a rotation matrix enhances lesion detection, promising improved diagnostic accuracy, and supporting treatment decisions for lumbar spinal stenosis patients.

Objective:To analyze the changes in plasma endothelin-1 (ET-1) concentrations in the patients with acute respiratory distress syndrome (ARDS).Methods:Fourteen patients with ARDS induced by trauma, 8 males and 6 females, aged 19-80 yr, were studied. The severity of ARDS was graded according to the Berlin definition of ARDS after admission to intensive care unit (ICU). Venous blood samples were obtained on 1st, 3rd and 5th days after admission to ICU, the plasma ET-1 concentrations were measured by radioimmunoassay, the pulmonary vascular permeability index (PVPI) was determined by PiCCO technique, and multiple organ dysfunction (MOD) score and lung injury score (LIS) were assessed. Spearman correlation of plasma ET-1 concentrations with MOD score, LIS and PVPI was analyzed.Results:MOD score, LIS, PVPI and plasma ET-1 concentrations were significantly decreased in mild ARDS patients ( n=5) as compared with moderate ARDS patients ( n=9, P<0.05). The plasma ET-1 concentration was positively correlated with MOD score, LIS and PVPI ( r=0.69, 0.76, 0.62, P<0.001). Conclusions:Plasma ET-1 concentrations can reflect the pulmonary vascular permeability and even the severity of the disease in the early stage of ARDS, so it is necessary to carry out dynamic monitoring in the patients.

Objective: To compare the clinical characteristics of different types of human adenovirus (HAdV) infection in hospitalized children with acute respiratory infection in Beijing, and to clarify the clinical necessity of adenovirus typing. Methods: In a cross-sectional study, 9 022 respiratory tract specimens collected from hospitalized children with acute respiratory infection from November 2017 to October 2019 in Affiliated Children's Hospital, Capital Institute of Pediatrics were screened for HAdV by direct immunofluorescence (DFA) and (or) nucleic acid detection. Then the Penton base, Hexon and Fiber gene of HAdV were amplified from HAdV positive specimens to confirm their HAdV types by phylogenetic tree construction. Clinical data such as laboratory results and imaging data were analyzed for children with predominate type HAdV infection using t, U, or χ² test. Results: There were 392 cases (4.34%) positive for HAdV among 9 022 specimens from hospitalized children with acute respiratory infection. Among those 205 cases who were successfully typed, 131 were male and 74 were female, age of 22.6 (6.7, 52.5) months,102 cases (49.76%) were positive for HAdV-3 and 86 cases (41.95%), HAdV-7, respectively, while 17 cases were confirmed as HAdV-1, 2, 4, 6, 14 or 21. In comparison of clinical characteristics between the predominate HAdV type 7 and 3 infection, significant differences were shown in proportions of children with wheezing (10 cases (11.63%) vs. 25 cases (24.51%)), white blood cell count >15 ×10⁹/L (4 cases (4.65%) vs.14 cases (13.73%)), white blood cell count <5×10⁹/L (26 cases (30.23%) vs.11 cases (10.78%)), procalcitonin level>0.5 mg/L (43 cases (50.00%) vs. 29 cases (28.43%)), multilobar infiltration (45 cases (52.33%) vs.38 cases (37.25%)), pleural effusion (23 cases (26.74%) vs. 10 cases (9.80%)), and severe adenovirus pneumonia (7 cases (8.14%) vs. 2 cases (1.96%)) with χ²=5.11, 4.44, 11.16, 9.19, 4.30, 9.25, 3.91 and P=0.024, 0.035, 0.001, 0.002, 0.038, 0.002, 0.048, respectively, and also in length of hospital stay (11 (8, 15) vs. 7 (5, 13) d, Z=3.73, P<0.001). Conclusions: HAdV-3 and 7 were the predominate types of HAdV infection in hospitalized children with acute respiratory tract infection in Beijing. Compared with HAdV-3 infection, HAdV-7 infection caused more obvious inflammatory reaction, more severe pulmonary symptoms, longer length of hospital stay, suggesting the clinical necessity of further typing of HAdVs.
Adenovirus Infections, Human/epidemiology* ; Adenoviruses, Human/genetics* ; Beijing/epidemiology* ; Child ; Child, Hospitalized ; Cross-Sectional Studies ; Female ; Humans ; Infant ; Male ; Phylogeny ; Respiratory Tract Infections/epidemiology*

Objective: To investigate the effects of continuing exercise and load-bearing interval exercise on skeletal muscle tissue cell morphology, Ras-related proteins 5 (Rab5) mRNA and protein expression and glucose metabolism in skeletal muscle of type 2 diabetic mellitus (T2DM) rats. Methods: Eight SD rats were selected as controls group (CR), the others SD rats were fed with high fat and high sugar diet for 6 weeks before injecting STZ (35 mg/kg) to construct the T2DM model. Twenty-four T2DM rats were randomly devided into T2DM model group (DRM), continuing exercise group (DCRE) and load-bearing interval exercise group (DWRE), 8 rats in each group. DCRE exercise protocol, that was 15 m/min (10 min), 20 m/min (40 min), 15 m/min (10 min), during the first 1～2 weeks, and 18 m/min (10 min), 25 m/min (40 min), 15 m/min (10 min), during the second 3～8 weeks. DWRE exercise protocol: load weight 15% / 1～2 weeks, 30% / 3～4 weeks, 45% / 5～8 weeks, with 15 m/min (5 min), 12 groups and 3 min rest between groups. After 8 weeks, pathological and morphological changes of skeletal muscle were observed by HE. Rab5 and Glucose transporte 4 (GLUT4) mRNA expressions of skeletal muscle were tested by qRT-PCR. Rab5 protein expression in skeletal muscle was tested by immunofluorescence histochemistry and Western blot, and plasma Rab5 and Glycosylated Hemoglobin (GHb) concentrations were detected by ELISA. Results: Comparison with CR, DRM showed pathological damage of skeletal muscle, the expressions of Rab5 mRNA, protein and GLUT4 mRNA were all decreased in skeletal muscle (P＜0.01), the serum levels of Rab5 and GHb were both significantly elevated (P＜0.01). Comparison with DRM, both DCRE and DWRE significantly improved pathological damages of skeletal muscle, the expressions of Rab5 mRNA, protein and GLUT4 mRNA were all increased in skeletal muscle (P＜ 0.05, P＜0.01), the serum levels of Rab5 and GHb were decreased (P＜0.05, P＜0.01), and there was no statistical difference between DCRE and DWRE groups (P＞0.05). Conclusion: Two exercise modes can improve the pathological injury of skeletal muscle in type 2 diabetic rats, and enhance GLUT4 transport capacity by improving the expression of Rab5 gene and protein in skeletal muscle, and alleviate the imbalance of glucose metabolism homeostasis in skeletal muscle. However, there was no significant difference between the effects of two exercise modes on Rab5 protein and glucose metabolism in skeletal muscle.
Animals ; Diabetes Mellitus, Experimental/metabolism* ; Diabetes Mellitus, Type 2/metabolism* ; Glucose/metabolism* ; Glycated Hemoglobin ; Insulin ; Muscle, Skeletal/metabolism* ; Physical Conditioning, Animal/methods* ; RNA, Messenger/metabolism* ; Rats ; Rats, Sprague-Dawley ; rab5 GTP-Binding Proteins/metabolism*

In recent years, the real-world studies (RWS) have attracted extensive attention, and the real-world evidence (RWE) has been accepted to support the drug development in China and abroad. However, there is still a lack of standards for the evaluation of the quality of RWE. It is necessary to formulate a quality evaluation and reporting specification for RWE especially in traditional Chinese medicine (TCM). To this end, under the guidance of China Association of Chinese Medicine, the Quality Evaluation and Reporting Specification for Real-World Evidence of Traditional Chinese Medicine (QUERST) Group, including 24 experts (clinical epidemiologists, clinicians, pharmacologists, ethical reviewer and statisticians), was established to develop the specification. This specification contains the listing of classification of RWS design and RWE, the general principles and methods of RWE quality evaluation (26 tools or scales), 25 types of bias in RWS, the special considerations in evaluating the quality of RWE of TCM, and the 19 reporting standards of RWE. This specification aims to propose the quality evaluation principles and key points of RWE, and provide guidance for the proper use of RWE in the development of TCM new drugs.
Medicine, Chinese Traditional ; China

To construct a non-communicable disease system recommended by WHO, develop the key techniques and promote their applications, obtain the main health indicators and understand the prevalence of chronic diseases, and provide support for the prevention, control and research of chronic diseases. Based on factor analysis, K-means clustering and multi-cluster random sampling, 30 typical sampling areas at provincial level were designed and constructed; By referring to WHO's Non-communicable Disease Surveillance Framework and the American behavioral risk factor sampling and questionnaire and combined with China's actual needs, a comprehensive surveillance system for chronic diseases, covering morbidity and mortality, risk factor exposure and community management and control of chronic diseases, was established, a "5+12+1" quality control system for surveillance data collection, management, analysis and feedback was formed and a three-level surveillance information management platform and information technology construction standards in the province were established, resulting the integration of life registration, chronic disease case reporting and community chronic disease management. Using these key techniques, we have obtained high-quality surveillance data of the whole province, produced the main health indicators, carried out research of chronic diseases, and analyze the prevalence and changing trend of the main chronic diseases and related risk factors to boost the government's practical projects for the reform of the people's livelihood and facilitate the construction of "Healthy Zhejiang". The successful experiences and key techniques have been applied in the construction of chronic disease surveillance system in some provinces in China.
China/epidemiology* ; Chronic Disease ; Chronic Disease Indicators ; Humans ; Noncommunicable Diseases ; Prevalence