ObjectiveTo construct a large language model （LLM）-based intelligent pre-consultation system for traditional Chinese medicine （TCM） to improve efficacy of clinical practice. MethodsA TCM large language model was fine-tuned using DeepSpeed ZeRO-3 distributed training strategy based on YAYI 2-30B. A weighted undirected graph network was designed and an agent-based syndrome differentiation model was established based on relationship data extracted from TCM literature and clinical records. An agent collaboration framework was developed to integrate the TCM LLM with the syndrome differentiation model. Model performance was comprehensively evaluated by Loss function， BLEU-4， and ROUGE-L metrics， through which training convergence， text generation quality， and language understanding capability were assessed. Professional knowledge test sets were developed to evaluate system proficiency in TCM physician licensure content， TCM pharmacist licensure content， TCM symptom terminology recognition， and meridian identification. Clinical tests were conducted to compare the system with attending physicians in terms of diagnostic accuracy， consultation rounds， and consultation duration. ResultsAfter 100 000 iterations， the training loss value was gradually stabilized at about 0.7±0.08， indicating that the TCM-LLM has been trained and has good generalization ability. The TCM-LLM scored 0.38 in BLEU-4 and 0.62 in ROUGE-L， suggesting that its natural language processing ability meets the standard. We obtained 2715 symptom terms， 505 relationships between diseases and syndromes， 1011 relationships between diseases and main symptoms， and 1 303 600 relationships among different symptoms， and constructed the Agent of syndrome differentiation model. The accuracy rates in the simulated tests for TCM practitioners， licensed pharmacists of Chinese materia medica， recognition of TCM symptom terminology， and meridian recognition were 94.09%， 78.00%， 87.50%， and 68.80%， respectively. In clinical tests， the syndrome differentiation accuracy of the system reached 88.33%， with fewer consultation rounds and shorter consultation time compared to the attending physicians （P＜0.01）， suggesting that the system has a certain pre- consultation ability. ConclusionThe LLM-based intelligent TCM pre-diagnosis system could simulate diagnostic thinking of TCM physicians to a certain extent. After understanding the patients' natural language， it collects all the patient's symptom through guided questioning， thereby enhancing the diagnostic and treatment efficiency of physicians as well as the consultation experience of the patients.

Objective To evaluate the performance of magnetic resonance imaging(MRI)multi-sequence feature imputation and fusion mutual model based on sequence deletion in differentiating high-grade glioma(HGG)from low-grade glioma(LGG).Methods We retrospectively collected multi-sequence MR images from 305 glioma patients,including 189 HGG patients and 116 LGG patients.The region of interest(ROI)of T1-weighted images(T1WI),T2-weighted images(T2WI),T2 fluid attenuated inversion recovery(T2_FLAIR)and post-contrast enhancement T1WI(CE_T1WI)were delineated to extract the radiomics features.A mutual-aid model of MRI multi-sequence feature imputation and fusion based on sequence deletion was used for imputation and fusion of the feature matrix with missing data.The discriminative ability of the model was evaluated using 5-fold cross-validation method and by assessing the accuracy,balanced accuracy,area under the ROC curve(AUC),specificity,and sensitivity.The proposed model was quantitatively compared with other non-holonomic multimodal classification models for discriminating HGG and LGG.Class separability experiments were performed on the latent features learned by the proposed feature imputation and fusion methods to observe the classification effect of the samples in two-dimensional plane.Convergence experiments were used to verify the feasibility of the model.Results For differentiation of HGG from LGG with a missing rate of 10%,the proposed model achieved accuracy,balanced accuracy,AUC,specificity,and sensitivity of 0.777,0.768,0.826,0.754 and 0.780,respectively.The fused latent features showed excellent performance in the class separability experiment,and the algorithm could be iterated to convergence with superior classification performance over other methods at the missing rates of 30%and 50%.Conclusion The proposed model has excellent performance in classification task of HGG and LGG and outperforms other non-holonomic multimodal classification models,demonstrating its potential for efficient processing of non-holonomic multimodal data.

Objective To evaluate the performance of magnetic resonance imaging(MRI)multi-sequence feature imputation and fusion mutual model based on sequence deletion in differentiating high-grade glioma(HGG)from low-grade glioma(LGG).Methods We retrospectively collected multi-sequence MR images from 305 glioma patients,including 189 HGG patients and 116 LGG patients.The region of interest(ROI)of T1-weighted images(T1WI),T2-weighted images(T2WI),T2 fluid attenuated inversion recovery(T2_FLAIR)and post-contrast enhancement T1WI(CE_T1WI)were delineated to extract the radiomics features.A mutual-aid model of MRI multi-sequence feature imputation and fusion based on sequence deletion was used for imputation and fusion of the feature matrix with missing data.The discriminative ability of the model was evaluated using 5-fold cross-validation method and by assessing the accuracy,balanced accuracy,area under the ROC curve(AUC),specificity,and sensitivity.The proposed model was quantitatively compared with other non-holonomic multimodal classification models for discriminating HGG and LGG.Class separability experiments were performed on the latent features learned by the proposed feature imputation and fusion methods to observe the classification effect of the samples in two-dimensional plane.Convergence experiments were used to verify the feasibility of the model.Results For differentiation of HGG from LGG with a missing rate of 10%,the proposed model achieved accuracy,balanced accuracy,AUC,specificity,and sensitivity of 0.777,0.768,0.826,0.754 and 0.780,respectively.The fused latent features showed excellent performance in the class separability experiment,and the algorithm could be iterated to convergence with superior classification performance over other methods at the missing rates of 30%and 50%.Conclusion The proposed model has excellent performance in classification task of HGG and LGG and outperforms other non-holonomic multimodal classification models,demonstrating its potential for efficient processing of non-holonomic multimodal data.

Abstract Oral diseases are the most important public health problems that have not received due attention, especially among college students. Based on the analysis of literatures on oral health status and behaviors of college students, the paper systematically summarizes the relevant research situation, points out the existing problems. Suggestions and countermeasures, such as attaching importance to the investigation and research on oral health of college students, improving the quality and level of scientific research on oral health of college students, exploring more effective comprehensive intervention measures for oral diseases and oral health education, should be implemented as projects are put forward, so as to provide basis for improving the oral health status of Chinese college students and promoting relevant research level.

Objective:To explore the risk factors and outcomes of central nervous system(CNS)complications associated with hematopoietic stem cell transplantation(HSCT).Methods:A total of 550 recipient after HSCT in the department of hematology of Tongji Hospital affiliated to Tongji Medical College of Huazhong University of Science and Technology from January 1 2019 to August 31 2021were enrolled.According to the occurrence of CNS complications, they were divided into the CNS group(24 cases)and the non CNS group(526 cases). The clinical information and prognosis were compared.We further analyzed the risk factors associated with CNS complications, and conducted multivariate logistic regression on statistically significant indicators.Cox regression analysis is conducted on prognostic factors such as age, gender and risk degree.Results:A total of 550 recipients were enrolled, of which 330 underwent allo-HSCT, and others received auto-HSCT.A total of 24 cases (4.36%)had CNS complications, of which 4 cases had 2 types of CNS complications.The type of CNS complications included intracranial infection(8 cases, 28.57%), transplantation-associated thrombotic microangiopathy(TA-TMA)(6 cases, 21.43%), central tumor invasion(4 cases, 14.29%), intracranial hemorrhage(4 cases, 14.29%), leucodystrophy(2 cases, 7.14%)and unexplained encephalopathy(4 cases, 14.29%). Logistic regression analysis of risk factors related to CNS complications showed that, Platelet implantation time( β=0.084, OR=1.088, P=0.048), CMV infection( β=1.295, OR=3.65, P=0.008)is positively correlated with the occurrence of CNS complications in HSCT recipients but age( β=-0.052, OR=0.949, P=0.004)is negatively correlated with it.Nine of the 24 cases(37.50%)who experienced CNS complications died, including 3 cases of intracranial infection, 3 cases of cerebral hemorrhage, 2 cases of TMA, and 1 case of unexplained encephalopathy.Platelet implantation time is an independent risk factor for poor prognosis of CNS complications in HSCT recipients. Conclusions:Our results indicated that, age, CMV infection and platelet implantation time were associated with the occurrence of CNS complications after HSCT.Platelet implantation time is an independent risk factor for poor prognosis of CNS complications in HSCT recipients.

Eyelid and surrounding tissue injury caused by trauma, burn, infection and other reasons will lead to different degrees of eyelid ectropion and incomplete eyelid closure after scar formation. At present, surgical treatment is the main treatment for cicatricial ectropion, but there is still the risk of scar formation again after operation. We used early ion beam drug introduction to treat 1 case of cicatricial ectropion after trauma, and achieved satisfactory result.

Eyelid and surrounding tissue injury caused by trauma, burn, infection and other reasons will lead to different degrees of eyelid ectropion and incomplete eyelid closure after scar formation. At present, surgical treatment is the main treatment for cicatricial ectropion, but there is still the risk of scar formation again after operation. We used early ion beam drug introduction to treat 1 case of cicatricial ectropion after trauma, and achieved satisfactory result.

Objective: To explore the incidence and influencing factors of child emotional neglect, and to provide evidence for the relevant departments to prevent child emotional neglect. Methods: Stratified random cluster sampling method was used to select 1 354 parents from a county of Henan Province. Anonymous questionnaire survey was conducted among them to investigate child emotional neglect, logistic regression analysis was used to analyze the influencing factors of emotional neglect. Results: The total incidence of child emotional neglect was 15.7%, and the rate of boys and girls was 14.7% and 16.8%, respectively, that of fathers and mothers was 18.3% and 15.1%, respectively. There were no statistically significant differences in rates between fathers and mothers’ emotional neglect behaviors, neither between boys and girls (χ2=1.18，3.14, P>0.05).90.2% parents could often/always console children when they were hurt or feared. Multivariate logistic regression analysis showed that risk factors of high emotional neglect incidence included living in rural areas, poorer family economic status, parents with lower social support and childhood neglect experience, biological parents are not married(OR=2.71, 2.56, 2.64, 2.66, 4.11, 1.76, P<0.05). Conclusion Parents’ emotional neglect of kindergarten children existed in the research area, attention should be paid to the emotional neglect of parents from rural areas, and parental education activities should be carried out to enhance parents’ parenting skills.

Objective To observe the short- and long-term efficacies of acupuncture in treating metabolic syndrome (MS). Method A hundred MS patients were randomized into an acupuncture group and a Western medication group. In addition to basic treatment, acupuncture was used in the acupuncture group, while Western medication was adopted in the Western medication group. The fasting plasma glucose (FPG), high-density lipoprotein (HDL), insulin (INS), triglyceride (TG), blood pressure (BP), weight circumference (WC), body mass index (BMI), and MS score were evaluated before and after treatment, and a month and 3 months after the treatment. Result There was no significant difference in comparing the short-term therapeutic efficacy between the two groups after treatment, while the long-term therapeutic efficacy of the acupuncture group was superior to that of the Western medication group (P<0.01); the FPG, INS, TG, HDL, BP, and MS score were significantly improved after treatment and a month after the treatment in both groups (P<0.01); the acupuncture group had significant improvements in WC and BMI after treatment and a month after the treatment (P<0.01), significantly superior to the Western medication group (P<0.01); the acupuncture group showed marked improvement in INS, HDL, WC, and BMI 3 months after the treatment (P<0.01), significantly superior to the Western medication group (P<0.01,P<0.05). Conclusion Acupuncture can produce a similar short-term but a better long-term therapeutic efficacy compared to Western medication. Hence, acupuncture can be taken as an assistant intervention for MS.

BACKGROUND:Pedicle screw fixation has been used in children with thoracic injury. Conventional method is screw implantation by hand. This method can meet the needs of surgery, but its accuracy was low, and incidence of complications was high. The application of digital image navigation aid module is possible. OBJECTIVE:To study accuracy and safety of digital image navigation aid module in thoracic pedicle screw placement. METHODS:Eight thoracic vertebral bodies were equal y and randomly assigned to the manual insertion group and the digital image navigation aid module group. Manual insertion group received manual screw insertion. In the digital image navigation aid module group, navigation aid module was made according to CT scan results combined with principle of reverse engineering and rapid prototyping. Pedicle screw was placed using the digital image navigation aid module. RESULTS AND CONCLUSION:(1) Success rate of once screw set was significantly higher in the digital imaging navigation aid module group than in the manual insertion group (P<0.05). (2) Twenty-eight screws were implanted in the digital image navigation aid module group and manual insertion group separately. The excellent and good rate of screw placement was 96%in the digital image navigation aid module group and 75%in the manual insertion group (P<0.05). (3) These findings suggested that digital image navigation aid module can effectively improve the success rate of pedicle screw insertion. Moreover, this method is simple, easy to operate, and can make a personalized nail placement program for each child.