Objective to establish a combined diagnosis model of scleroderma related genes based on gene expression comprehensive database(GEO)and artificial neural network(ANN)and to evaluate its effect and to predict and analyze targeted traditional Chinese medicine.Methods two scleroderma chips GSE23741 and GSE95065 were obtained from the GEO database as the training group data set.Random forest and lasso regression algorithms were used to screen the key genes of scleroderma and construct the ANN model for the diagnosis of scleroderma.The validation data sets GSE76807,GSE32413 and GSE59785 were used to verify the model,and the area under curve(AUC)analysis was used to evaluate the clinical application value of ANN model.The relative expression of key gene mRNA was verified by RT-qPCR experiment.The CIBERSORT algorithm was used to estimate the bioinformatics association between scleroderma and the screened biomarkers.Finally,the key genes were used to screen the targeted traditional Chinese medicine.Results A total of 167 differential genes were obtained.Furthermore,the five most relevant key genes(SERPINE2,SFRP4,SUGCT,FBLN5,NRXN2)were screened by machine learning,and the artificial neural network diagnosis model was constructed.The model was used to draw the subject operating characteristic(ROC)curves diagnosed by the training group and the verification group,and the AUC value of the training group was 1.000.The AUC of verification group were 0.770,0.795 and 0.872 respectively.The result of RT-qPCR experiment is consistent with that of machine learning algorithm.Immune cell infiltration analysis showed that the relative content of memory CD4+T cells was significantly increased in scleroderma group,while the relative content of γ δ T cells in normal group was significantly increased.Key genes are associated with macrophage M1,T cells,memory activated CD4+T cells,resting mast cells,CD8+T cells and so on.According to the key genes,12 traditional Chinese medicines were screened.Most of the four qi and five flavors belong to warm,cold,flat,sweet,pungent and bitter,and most of them belong to the meridians of liver,spleen and lung.Conclusion the artificial neural network diagnosis model of key genes of scleroderma is constructed,which can be used in clinical diagnosis of scleroderma,and the potential targeted traditional Chinese medicine for the treatment of scleroderma is predicted,which provides a new perspective for exploring the pathogenesis of scleroderma.

Objective To observe the value of CT radiomics for differentiating spinal bone islands(BI)and osteoblastic metastases(OBM).Methods Data of 109 BI lesions in 98 patients and 282 OBM lesions in 158 patients(including 103 OBM in 48 lung cancer cases,86 OBM in 52 breast cancer cases and 93 OBM in 58 prostate cancer cases)from 3 medical institutions were retrospectively analyzed.Data obtained from institution 1 were used as the internal dataset and divided into internal training set and internal validation set at a ratio of 7∶3,from institution 2 and 3 were used as external dataset.All datasets were divided into female data subset(including OBM of female lung cancer and breast cancer)and male data subset(including OBM of male lung cancer and prostate cancer).Radiomics features were extracted and screened to construct 3 different support vector machine(SVM)models,including model1 for distinguishing BI and OBM,model2 for differentiating OBM of female lung cancer and breast cancer,and model3 for differentiating OBM of male lung cancer and prostate cancer.Diagnostic efficacy of model1,CT value alone and 3 physicians(A,B,C)for distinguishing BI and OBM were assessed,as well as differentiating efficacy for different OBM of model2 and model3.Receiver operating characteristic(ROC)curves were drawn,and area under the curves(AUC)were calculated and compared.The differential diagnostic efficacy of model2 and model3 were also assessed with ROC analysis and AUC.Results AUC of model1 for distinguishing spinal OBM from BI in internal training set,internal validation set and external dataset was 0.99,0.98 and 0.86,respectively.In internal training set,model1 had higher AUC for distinguishing BI and OBM than that of physician A(AUC=0.78),B(AUC=0.87)and C(AUC=0.93)as well as that of mean CT value(AUC=0.78,all P＜0.05).AUC in internal training set,internal validation set and external dataset of model2 for identifying female lung cancer and breast cancer OBM was 0.79,0.75 and 0.73,respectively,of model3 for discriminating male lung cancer from prostate cancer OBM was 0.77,0.74 and 0.77,respectively.Conclusion CT radiomics SVM model might reliablely distinguish OBM and BI.

Objective:To explore diagnostic value of artificial intelligence(AI)technique on intracranial aneurysm(IA)in multi-slices computed tomography angiography(CTA)for head and neck.Methods:A retrospective analysis was performed on the cases data of 137 patients with suspected IA who were firstly diagnosed in Fuzong Teaching Hospital of Fujian University of Traditional Chinese Medicine from January 2022 to December 2023,all of them underwent CTA and digital subtraction angiography(DSA)examinations.The results of DSA examination were taken as the"gold standard".The work efficiency and detection effectiveness of two groups for IA were analyzed through manual post-processing and diagnosis(manual group)and CTA with intelligent assistant diagnostic system on head and neck(AI group).Results:In 137 patients with suspected IA,the DSA examination showed that there were 139 aneurysms in 112 patients.The post-processing time was(190.25±30.71)s in AI group,which was significantly shorter than that[(954.37±95.68)s]in manual group,and the difference was statistically significant(t=92.020,P＜0.001),and the image quality of AI group was higher.In the manual group,102 cases with aneurysms were found,which showed good consistency with the"gold standard"(Kappa=0.713),while the AI group found 93 cases with aneurysms,which showed moderate consistency with the"gold standard"(Kappa=0.518).Both groups had higher diagnostic accuracy for the presence or absence of subarachnoid hemorrhage of intracranial aneurysm,but the difference of that between them was not statistically significant difference(P＞0.05).The detection rates of anterior cerebral artery,middle cerebral artery,posterior cerebral artery,anterior communicating artery,posterior communicating artery,vertebral-basilar artery,and artery aneurysms of internal carotid in manual group and AI group were respectively 80.00％vs.80.00％,92.59％vs.81.48％,100.00％vs.100.00％,84.00％vs.52.00％,83.33％vs.0％,80.00％vs.60.00％,69.23％vs.57.69％,and both of them did not found lesions of posterior inferior cerebellar artery.The overall detection rates of two groups were respectively 78.42％and 57.55％,and the overall detection rate of lesion locations of manual group was higher than that of AI group,and the difference in overall coincidence rate between two groups was statistically significant(x2=13.899,P＜0.05).DSA confirmed 7 lesions which aneurysm's diameter was＜3 mm,and the manual group and the AI group identified respectively 3 cases and 1 case for the 7 lesions.A total of 60 lesions with 3～5 mm diameter were found by DSA,and the manual group and the AI group identified respectively 50 cases and 43 cases for the 60 lesions.A total of 72 lesions with＞5 mm diameter were found by DSA,and the manual group and the AI group identified respectively 63 cases and 63 cases,without statistically significant difference between them(P＞0.05).Conclusion:AI technique can significantly improve work efficiency of diagnosis for intracranial aneurysm,and obtain better processing effects of image.Although AI technique has shown a certain value of reference and auxiliary in diagnosing intracranial aneurysms,its diagnostic accuracy and precise in identifying lesion location are still lower than those of manual diagnosis,and it need further deep learning to improve its diagnostic ability.

Objective:To investigate the effects of 2 650 MHz radiofrequency (RF) exposure on the behavior and neurotransmitter release of mice.Methods:Adult male C57BL/6N mice were divided into a normal control (CON) group and a radiofrequency radiation (RFR) group using the random number table method. The mice in the RFR group were subjected to single-dose whole-body exposure to a uniform 2 650 MHz RF electromagnetic field for 3 h. During the RF exposure, the field strength in the effective working area of the RF radiation platform was measured using an electromagnetic radiation analyzer, and the changes in the anal temperature of the mice were monitored using an optical fiber thermometer. Moreover, the changes in the cognition, social interaction, and emotion of the mice were determined through the new object recognition test, social preference test, and open field test. Finally, the changes in the hippocampal neurotransmitter release levels of the mice were detected using microdialysis sampling and mass spectrometry, and the changes in the hippocampal tissue structure and ultrastructure were observed via microscopy.Results:Under the test conditions, RF radiation improved the anal temperature of the mice, with a maximum increasing amplitude of 0.61℃, falling within the range of thermal safety. The mice in the RFR group experienced a significant decrease in the frequency and time for exploring new objects ( t=4.50, 2.53, P < 0.05) in the new object recognition test, a significant decrease in the frequency ( t=0.08, P<0.01) and time ( t=0.03, P<0.05) for exploring other mice in the social preference test, and no significant change in the frequency and time for exploring the central area ( P > 0.05) in the open field test. Compared to the CON group, the RFR group showed an increase in the release of 5-hydroxytryptamine (5-HT) ( t=-2.56, P < 0.05) and a decrease in the release of acetylcholine (ACh) ( t=2.21, P < 0.05), no significant difference in the release of glutamate (Glu) and γ-aminobutyric acid (GABA) ( P > 0.05), and no evident damage to the hippocampal tissue and structure and synaptic ultrastructure. Conclusions:2 650 MHz RF radiation may induce cognitive impairment and abnormal social preference in mice, which is attributed to neuronal dysfunctions and neurotransmitter release disorders under RF exposure.

Objective:To clarify the effects of microwave radiation on anxiety-like behavior, the histomorphology of the secondary visual cortex, and calcium activity in neurons.Methods:36 C57BL/6N mice were selected and divided into control group and microwave radiation group according to the random number table method. In the simple behavioral testing, there were 8 mice in the control group and 7 mice in the radiation group. Combining fiber optic recording with behavioral experiments, there were 8 mice in the control group and 7 mice in the radiation group. Hematoxylin-eosin (HE) staining was conducted with 3 mice in each group. A high-power microwave simulated source in the X-band with a center frequency of 9.875 GHz and an average power density of 12 mW/cm 2 was used to irradiate the mice for 15 minutes, establishing a microwave radiation animal model. Then, anxiety-like behavior changes in the radiation group were identified using the open-field and elevated plus maze (EPM) tests. The effects of microwave radiation on the histomorphology of the secondary visual cortex were investigated using HE staining and optical microscopy. Based on the genetically encoded calcium imaging technique, as well as optical fiber recording combined with behavioral paradigms in the open field and the EPM, the changes of calcium activity in neurons in the V2M region of the secondary visual cortex were detected. Results:Compared to the control group, the radiation group showed a significant decrease in the frequency of exploring the central region of the open field and the open arm of the EPM ( t = 2.24, 3.10, P < 0.05). Furthermore, the radiation group exhibited the degeneration and apoptosis of some neurons in the secondary visual cortex, primarily manifested as pyknosis and deep staining, cell body shrinkage, and the slightly widening of perivascular space. Fiber optic recordings and behavioral experiments indicated that compared to the control group, mice in the radiation group exhibited significantly increased calcium activities in neurons of the secondary visual cortex when exploring the central region of the open field ( t = -2.75, P < 0.05) or the open arm of the EPM ( t = -2.77, -3.41, P < 0.05) compared to those before radiation after microwave exposure. Conclusions:Microwave radiation can induce anxiety-like behaviors and histopathological changes in the secondary visual cortex. Increased calcium activity in neurons of the secondary visual cortex is proved to be an important mechanism underlying the changes in anxiety-like behavior due to microwave radiation.

Infectious disease is a major public health problem threatening human health and social development, and infectious disease teaching is an important part of public health education. However, traditional infectious disease teaching faces challenges such as overly theoretical and fragmented content, overly traditional and monotonous methods, and a narrow and isolated vision, which makes it difficult to adapt to the current complex and changing situation of infectious disease prevention and control. The “One Health” concept emphasizes interdisciplinary, cross-sectoral, and cross-regional communication and cooperation to achieve the harmonious unity of health for humans, animals, and the environment, which is significant in preventing and controlling infectious diseases. This paper proposes some ideas for reforming infectious disease teaching based on the One Health concept, which aims to improve the quality and effectiveness of infectious disease teaching by recognizing the close connection between human health, animal health, and environmental health. Specifically, it is suggested to reform the teaching in four aspects: introducing the basic concepts, principles, and practices of One Health, increasing the content of infectious diseases related to One Health, adopting diversified and interactive teaching methods, and establishing cross-disciplinary teaching cooperation. Concrete recommendations are provided for each aspect. This paper argues that reforming infectious disease teaching based on the One Health concept is conducive to cultivating public health talents with global vision, systems thinking, and cross-disciplinary collaboration capabilities, providing robust talent support for coping with emerging, re-emerging, and endemic infectious disease threats.

AIM To study the phenylpropanoids from Brandisia hancei Hook.f.and their antioxidant activities.METHODS The extract from B.hancei was isolated and purified by Rp-C18,MCI,semi-preparative HPLC,silica gel and Sephadex LH-20,then the structures of obtained compounds were identified by physicochemical properties and spectral data.The cytotoxicities was determined by MTT method,and the antioxidant activities were determined by DPPH and ABTS+free radical scavenging methods.RESULTS Fifteen phenylpropanoids were isolated and identified as(+)-pinonesinol(1),(-)-medioresinol(2),(-)-syringaresinol(3),buddlenol D(4),(7R,7'R,7″S,8S,8'S,8″S)-4',5″-dihydroxy-3,5,3',4″-tetramethoxy-7,9':7',9-diepoxy-4,8″-oxy-8,8'-sesquineo-lignan-7″,9″-diol(5),(-)-(7R,7'R,7″R,8S,8'S,8″S)-4',4″-dihydroxy-3,3',3″,5-tetramethoxy-7,9':7',9-diepoxy-4,8″-oxy-8,8'-sesquineolignan-7″,9″-diol(6),hedyotol A(7),dracunculifoside R(8),acteoside(9),isoacteoside(10),arenarioside(11),isomartynoside(12),curcasinlignan B(13),erythro-2,3-bis(4-hydroxy-3-methoxyphenyl)-3-ethoxypropan-l-ol(14),citrusin C(15).Compounds 1-4 and 9-10 had no obvious cytotoxicity to HepG2 hepatoma cells.Compounds 1,3,9,10 and 12 had strong scavenging activities against DPPH radicals.Compounds 1-3,9-10,12 and 14 showed strong scavenging activities against ABTS+radical.CONCLUSION Compounds 1-8 and 12-15 are isolated from genus Brandisia for the first time.The phenylpropanoids from B.hancei show strong antioxidant activities.

Objective: To establish a comprehensive diagnostic classification model of lateral cephalograms based on artificial intelligence (AI) to provide reference for orthodontic diagnosis. Methods: A total of 2 894 lateral cephalograms were collected in Department of Orthodontics, Capital Medical University School of Stomatology from January 2015 to December 2021 to construct a data set, including 1 351 males and 1 543 females with a mean age of (26.4± 7.4) years. Firstly, 2 orthodontists (with 5 and 8 years of orthodontic experience, respectively) performed manual annotation and calculated measurement for primary classification, and then 2 senior orthodontists (with more than 20 years of orthodontic experience) verified the 8 diagnostic classifications including skeletal and dental indices. The data were randomly divided into training, validation, and test sets in the ratio of 7∶2∶1. The open source DenseNet121 was used to construct the model. The performance of the model was evaluated by classification accuracy, precision rate, sensitivity, specificity and area under the curve (AUC). Visualization of model regions of interest through class activation heatmaps. Results: The automatic classification model of lateral cephalograms was successfully established. It took 0.012 s on average to make 8 diagnoses on a lateral cephalogram. The accuracy of 5 classifications was 80%-90%, including sagittal and vertical skeletal facial pattern, mandibular growth, inclination of upper incisors, and protrusion of lower incisors. The acuracy rate of 3 classifications was 70%-80%, including maxillary growth, inclination of lower incisors and protrusion of upper incisors. The average AUC of each classification was ≥0.90. The class activation heat map of successfully classified lateral cephalograms showed that the AI model activation regions were distributed in the relevant structural regions. Conclusions: In this study, an automatic classification model for lateral cephalograms was established based on the DenseNet121 to achieve rapid classification of eight commonly used clinical diagnostic items.
Male ; Female ; Humans ; Young Adult ; Adult ; Artificial Intelligence ; Deep Learning ; Cephalometry ; Maxilla ; Mandible/diagnostic imaging*

Objective: To develop a multi-classification orthodontic image recognition system using the SqueezeNet deep learning model for automatic classification of orthodontic image data. Methods: A total of 35 000 clinical orthodontic images were collected in the Department of Orthodontics, Capital Medical University School of Stomatology, from October to November 2020 and June to July 2021. The images were from 490 orthodontic patients with a male-to-female ratio of 49∶51 and the age range of 4 to 45 years. After data cleaning based on inclusion and exclusion criteria, the final image dataset included 17 453 face images (frontal, smiling, 90° right, 90° left, 45° right, and 45° left), 8 026 intraoral images [frontal occlusion, right occlusion, left occlusion, upper occlusal view (original and flipped), lower occlusal view (original and flipped) and coverage of occlusal relationship], 4 115 X-ray images [lateral skull X-ray from the left side, lateral skull X-ray from the right side, frontal skull X-ray, cone-beam CT (CBCT), and wrist bone X-ray] and 684 other non-orthodontic images. A labeling team composed of orthodontic doctoral students, associate professors, and professors used image labeling tools to classify the orthodontic images into 20 categories, including 6 face image categories, 8 intraoral image categories, 5 X-ray image categories, and other images. The data for each label were randomly divided into training, validation, and testing sets in an 8∶1∶1 ratio using the random function in the Python programming language. The improved SqueezeNet deep learning model was used for training, and 13 000 natural images from the ImageNet open-source dataset were used as additional non-orthodontic images for algorithm optimization of anomaly data processing. A multi-classification orthodontic image recognition system based on deep learning models was constructed. The accuracy of the orthodontic image classification was evaluated using precision, recall, F1 score, and confusion matrix based on the prediction results of the test set. The reliability of the model's image classification judgment logic was verified using the gradient-weighted class activation mapping (Grad-CAM) method to generate heat maps. Results: After data cleaning and labeling, a total of 30 278 orthodontic images were included in the dataset. The test set classification results showed that the precision, recall, and F1 scores of most classification labels were 100%, with only 5 misclassified images out of 3 047, resulting in a system accuracy of 99.84%(3 042/3 047). The precision of anomaly data processing was 100% (10 500/10 500). The heat map showed that the judgment basis of the SqueezeNet deep learning model in the image classification process was basically consistent with that of humans. Conclusions: This study developed a multi-classification orthodontic image recognition system for automatic classification of 20 types of orthodontic images based on the improved SqueezeNet deep learning model. The system exhibitted good accuracy in orthodontic image classification.
Humans ; Male ; Female ; Child, Preschool ; Child ; Adolescent ; Young Adult ; Adult ; Middle Aged ; Deep Learning ; Reproducibility of Results ; Radiography ; Algorithms ; Cone-Beam Computed Tomography

Objective: To investigate the timing of pericardial drainage catheter removal and restart of the anticoagulation in patients with atrial fibrillation (AF) suffered from perioperative pericardial tamponade during atrial fibrillation catheter ablation and uninterrupted dabigatran. Methods: A total of 20 patients with pericardial tamponade, who underwent AF catheter ablation with uninterrupted dabigatran in Beijing Anzhen Hospital from January 2019 to August 2021, were included in this retrospective analysis. The clinical characteristics of enrolled patients, information of catheter ablation procedures, pericardial tamponade management, perioperative complications, the timing of pericardial drainage catheter removal and restart of anticoagulation were analyzed. Results: All patients underwent pericardiocentesis and pericardial effusion drainage was successful in all patients. The average drainage volume was (427.8±527.4) ml. Seven cases were treated with idarucizumab, of which 1 patient received surgical repair. The average timing of pericardial drainage catheter removal and restart of anticoagulation in 19 patients without surgical repair was (1.4±0.7) and (0.8±0.4) days, respectively. No new bleeding, embolism and death were reported during hospitalization and within 30 days following hospital discharge. Time of removal of pericardial drainage catheter, restart of anticoagulation and hospital stay were similar between patients treated with idarucizumab or not. Conclusion: It is safe and reasonable to remove pericardial drainage catheter and restart anticoagulation as soon as possible during catheter ablation of atrial fibrillation with uninterrupted dabigatran independent of the idarucizumab use or not in case of confirmed hemostasis.
Humans ; Atrial Fibrillation/drug therapy* ; Dabigatran/therapeutic use* ; Cardiac Tamponade/complications* ; Anticoagulants/therapeutic use* ; Retrospective Studies ; Treatment Outcome ; Drainage/adverse effects* ; Catheter Ablation ; Catheters/adverse effects*