Objective:To develop a deep learning-based system for real-time recognition and classification of portal hypertensive gastropathy (PHG) and evaluate its ability to assist junior endoscopists.Methods:A total of 2 848 gastroscopy images from 832 patients with liver cirrhosis were selected from Digestive Endoscopy Center databases of Renmin Hospital of Wuhan University, Wuhan Hospital of Traditional Chinese and Western Medicine, and the Second Hospital of Jingzhou from January 2015 to October 2023. This system referred to 3 endoscopic features of Baveno Ⅱ scoring system. Three models were developed respectively for gastric antral vascular ectasia (GAVE), mosaic-like pattern (MLP), and red marks (RM). The specific classification references were as follows: (1) GAVE model: 0 no, 1 yes; (2) MLP model: 0 no, 1 mild, 2 severe; (3) RM model: 0 no, 1 isolated, 2 fused. The classification results for endoscopic characteristics of PHG of 3 endoscopy experts were taken as the gold standard. The yolov8-m model was used for training. The training dataset, validation dataset, and test dataset were allocated at a ratio of 8∶1∶1. The test dataset was used to evaluate the performance of models and their auxiliary effects on endoscopists. The accuracy, recall, precision, specificity and Kappa coefficient were calculated. Results:The accuracy, recall, specificity of GAVE model were 96.0% (48/50), 87.5% (7/8) and 97.6% (41/42). There was no significant difference between its accuracy and the gold standard ( χ2=316.226, P=1.000). The precision of GAVE1 and GAVE0 were 87.5% (7/8) and 97.6% (41/42) respectively. The accuracy of MLP model was 84.1% (132/157), and there was no significant difference compared with the gold standard ( χ2=3.286, P=0.193). The precision and recall of MLP2 were 88.2% (15/17) and 75.0% (15/20). The precision and recall of MLP1 were 77.9% (60/77) and 88.2% (60/68). The precision and recall of MLP0 were 90.5% (57/63) and 82.6% (57/69). The accuracy of RM model was 87.9% (123/140), and there was no significant difference compared with the gold standard ( χ2=2.891, P=0.409). The precision and recall of RM2 were 94.7% (18/19) and 78.3% (18/23). The precision and recall of RM1 were 72.2% (26/36) and 81.3% (26/32). The precision and recall of RM0 were 92.9% (79/85) and 92.9% (79/85). The mean accuracy of the three junior endoscopists, with and without the assistance of the GAVE model, MLP model, and RM model, respectively increased from 95.3% to 99.3%, from 83.9% to 91.9%, and from 81.9% to 83.1%. The overall consistency analysis of the 3 junior endoscopists with the gold standard indicated that the consistency of the GAVE model before and after assistance was extremely strong (both an overall Kappa of 1.000); the consistency before assistance of the MLP model was moderate (with an overall Kappa of 0.601), which increased to extremely strong after assistance (with an overall Kappa of 0.964); and the consistency of the RM model before and after assistance was also relatively strong (with an overall Kappa of 0.792 before and 0.798 after). Conclusion:The deep learning system accurately identifies and classifies PHG features and significantly enhances diagnostic performance of junior endoscopists.

Objective:To develop a deep learning-based system for real-time recognition and classification of portal hypertensive gastropathy (PHG) and evaluate its ability to assist junior endoscopists.Methods:A total of 2 848 gastroscopy images from 832 patients with liver cirrhosis were selected from Digestive Endoscopy Center databases of Renmin Hospital of Wuhan University, Wuhan Hospital of Traditional Chinese and Western Medicine, and the Second Hospital of Jingzhou from January 2015 to October 2023. This system referred to 3 endoscopic features of Baveno Ⅱ scoring system. Three models were developed respectively for gastric antral vascular ectasia (GAVE), mosaic-like pattern (MLP), and red marks (RM). The specific classification references were as follows: (1) GAVE model: 0 no, 1 yes; (2) MLP model: 0 no, 1 mild, 2 severe; (3) RM model: 0 no, 1 isolated, 2 fused. The classification results for endoscopic characteristics of PHG of 3 endoscopy experts were taken as the gold standard. The yolov8-m model was used for training. The training dataset, validation dataset, and test dataset were allocated at a ratio of 8∶1∶1. The test dataset was used to evaluate the performance of models and their auxiliary effects on endoscopists. The accuracy, recall, precision, specificity and Kappa coefficient were calculated. Results:The accuracy, recall, specificity of GAVE model were 96.0% (48/50), 87.5% (7/8) and 97.6% (41/42). There was no significant difference between its accuracy and the gold standard ( χ2=316.226, P=1.000). The precision of GAVE1 and GAVE0 were 87.5% (7/8) and 97.6% (41/42) respectively. The accuracy of MLP model was 84.1% (132/157), and there was no significant difference compared with the gold standard ( χ2=3.286, P=0.193). The precision and recall of MLP2 were 88.2% (15/17) and 75.0% (15/20). The precision and recall of MLP1 were 77.9% (60/77) and 88.2% (60/68). The precision and recall of MLP0 were 90.5% (57/63) and 82.6% (57/69). The accuracy of RM model was 87.9% (123/140), and there was no significant difference compared with the gold standard ( χ2=2.891, P=0.409). The precision and recall of RM2 were 94.7% (18/19) and 78.3% (18/23). The precision and recall of RM1 were 72.2% (26/36) and 81.3% (26/32). The precision and recall of RM0 were 92.9% (79/85) and 92.9% (79/85). The mean accuracy of the three junior endoscopists, with and without the assistance of the GAVE model, MLP model, and RM model, respectively increased from 95.3% to 99.3%, from 83.9% to 91.9%, and from 81.9% to 83.1%. The overall consistency analysis of the 3 junior endoscopists with the gold standard indicated that the consistency of the GAVE model before and after assistance was extremely strong (both an overall Kappa of 1.000); the consistency before assistance of the MLP model was moderate (with an overall Kappa of 0.601), which increased to extremely strong after assistance (with an overall Kappa of 0.964); and the consistency of the RM model before and after assistance was also relatively strong (with an overall Kappa of 0.792 before and 0.798 after). Conclusion:The deep learning system accurately identifies and classifies PHG features and significantly enhances diagnostic performance of junior endoscopists.

Objective:To establish a nomogram model for predicting the hyper-progression recurrence after hepatectomy in patients with hepatocellular carcinoma (HCC) based on circulating tumor cells (CTC).Methods:Clinical data of 231 HCC patients undergoing hepatectomy at the Department of Hepatobiliary Surgery, Guangxi Medical University Cancer Hospital from January 2013 to December 2022 were retrospectively analyzed, including 200 males and 31 females, aged 46(39, 52) years old. Patients were divided into two groups: the modeling group ( n=154) and the validation group ( n=77). According to the state of postoperative hyper-progression recurrence, patients in the modeling group were subdivided into hyper-progression recurrence ( n=39) and non-hyper-progression recurrence group ( n=115). Patients in the validation group were also subdivided into hyper-progression recurrence ( n=16) and non-hyper-progression recurrence group ( n=61). Clinicopathological data such as the total CTC count, alpha-fetoprotein, and postoperative pathology were collected. Logistic regression analysis was used to analyze the influencing factors of postoperative hyper-progression recurrence. A nomogram model was established based on the results of multivariate logistic regression analysis. The receiver operating characteristic (ROC) curve, calibration curve, decision curve analysis (DCA) and clinical impact curve (CIC) were used to validate the nomogram model. Results:Multivariate logistic regression analysis showed that HCC patients with age ≤45 years old ( OR=6.704, 95% CI: 1.619-27.760, P=0.009), incomplete tumor capsule ( OR=13.292, 95% CI: 3.084-57.295, P=0.001), high total numbers of CTC ( OR=1.101, 95% CI: 1.023-1.186, P=0.011) and high Ki67 index ( OR=52.659, 95% CI: 3.215-862.604, P=0.005) had a high risk of hyper-progression recurrence after hepatectomy. The above three preoperative variables were integrated to construct a nomogram model. The calibration curve showed that the predicted results of the nomogram model were in good agreement with the actual results. The ROC curves of the nomogram model for predicting hyper-progression recurrence after hepatectomy in HCC patients were plotted, and the area under the curve was 0.907 (95% CI: 0.856-0.959) and 0.833 (95% CI: 0.721-0.945) in the modeling group and validation group, respectively. DCA showed that the nomogram model could be used as a valuable predictive tool for the hyper-progression recurrence after hepatectomy. The CIC showed that the population judged by the nomogram model was highly matched with the actual population with hyper-progression recurrence. Conclusions:This study established a nomogram model based on age, tumor capsular integrity and total CTC count, which could accurately predict the postoperative hyper-progression recurrence in HCC patients before hepatectomy. The model is promising in guiding clinical practice after further validation.

Objective:To evaluate the necessity, safety and efficacy of endovascular treatment for cerebral infarction caused by middle cerebral artery (MCA) stenosis with hypoperfusion in the blood supply area of the lenticulostriate artery.Methods:The clinical and surgical data of patients with MCA atherosclerotic disease who underwent endovascular treatment in the First Affiliated Hospital of Zhengzhou University from January 2014 to October 2021 were retrospectively analyzed. A total of 6 patients with cerebral infarction caused by MCA stenosis with hypoperfusion in the blood supply area of the lenticulostriate artery were selected. The preoperative and postoperative clinical imaging characteristics, perioperative complications and follow-up of these 6 patients were summarized and evaluated.Results:After the endovascular treatment, the imaging of the lenticulostriate artery in all the 6 patients was clearer than that before the operation, and the number of main trunks of the lenticulostriate artery shown by imaging in 2 patients was more than that before operation. The computer tomography perfusion of 6 patients after the endovascular treatment showed that perfusion in the supply area of the lenticulostriate artery was significantly improved compared with pre-operation. No stroke, transient ischemic attack (TIA) and death occurred during the perioperative period. The time of clinical follow-up was 360 (322, 495) days, and there were no stroke, TIA or death occurring in the corresponding artery. All the 6 patients underwent imaging follow-up, of which 3 patients underwent digital subtraction angiography and 3 underwent CT angiography. The lumen of the target vessels showed patency in all patients.Conclusions:With rigorous imaging evaluation, endovascular treatment may be safe and effective for cerebral infarction caused by MCA stenosis with hypoperfusion in the blood supply area of the lenticulostriate artery.

Objective:To explore the effect of key amino acid mutations of 3D protein in enterovirus 71 on viral proliferation, and infer the mechanism of mutation affecting viral proliferation according to the tertiary structure model of 3D protein.Methods:The proliferation characteristics of the mutant strain that was constructed by site-directed mutagenesis and reverse genetics using the pMD19T-SDLY107-EGFP constructed from the fatal strain SDLY107 as a template were measured. Meanwhile the tertiary structure of the 3D protein was predicted, and then the possible mechanism of the mutation affecting the proliferation ability of the virus was speculated according to the functional characteristics of the domain of the 3D protein.Results:Two mutant viruses, eGFP-EV-A71 (S37N) and eGFP-EV-A71 (R142K), were successfully constructed by double enzyme digestion and viral gene sequencing. The proliferation rate of the two mutant strains in RD cells was significantly lower than that of the parent strains. The 3D protein tertiary structure prediction model showed that the 3D protein consisted of three domains: "Finger" , "thumb" and "Palm" , constituting a cupped right-handed structure. S37N and R142K are located in the "thumb" domain and " finger" domain, respectively. The "thumb" and "finger" domains have important effects on the activity of 3D polymerase and the stability of protein.Conclusions:Mutations at S37N and R142K sites of EV71 3D protein decrease the replication ability of EV71, and these two mutations may affect the proliferation of EV71 virus by changing the 3D protein polymerase activity and the interactions between multiple domains.

Objective:To analyze the risk factors and human papillomavirus (HPV) genotypes distribution in patients with condyloma acuminatum (CA) in two regions of Shandong province.Methods:From August 2019 to December 2020, an anonymous questionnaire survey of CA patients was conducted in three hospitals in Jinan City and Jining City, Shandong Province, and samples were collected for HPV typing. Multivariate binary logistic regression was used to analyze the risk factors of CA recurrence. HPV typing was detected by PCR-reverse dot blot hybridization.Results:A total of 653 questionnaires were collected, and the valid questionnaires accounted for 98.77% (645/653). Recurrence of the disease occurred in 174 patients, with a recurrence rate of 26.98%. Univariate analysis showed that there were statistically significant differences in the distribution of CA recurrence among residence time at current address, sexual frequency, genitalia cleaning, and knowledge of preventing HPV infection ( P<0.05). Multivariate binary logistic regression showed that knowing how to prevent HPV infection was a significant factor that influences CA recurrence. A total of 428 patients underwent HPV typing, and the positive detection rate of HPV was 98.60% (422/428). The top three positive rates were HPV6 (57.58%), HPV11 (36.49%) and HPV16 (11.37%). The main type of infection was low-risk HPV, accounting for 51.42% (217/422). Conclusions:CA patients have the phenomenon of "separation of knowledge and action" , so it is necessary to strengthen health education and behavioral intervention, guide the population to correctly treat sexual behavior, and improve self-prevention awareness and risk awareness.

Objective:We try to screen out predictive indicators with higher value by analyzing the differences in clinical and laboratory indicators between severe fever with thrombocytopenia syndrome (SFTS) patients in the intensive care unit (ICU) group and non-ICU group.Methods:The clinical and laboratory index data of 69 SFTS patients diagnosed in the laboratory in a hospital from June to December 2019 were retrospectively collected. According to the clinical outcome of the patients, they were divided into ICU and non-ICU groups. The differences in clinical manifestations and laboratory indicators between the two groups were analyzed. The receiver operating characteristic curve (ROC) was used to screen the more valuable predictive indicators.Results:Compared with the non-ICU group, ICU group SFTS patients had significantly higher procalcitonin (PCT), C-reactive protein (CRP), alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), glutamyl transpeptidase (GGT), leucine aminopeptidase (LAP), glutamate dehydrogenase (GDH), adenosine deaminase (ADA), cystatin C (Cys C), α-hydroxybutyrate dehydrogenase (α-HBDH), creatine kinase (CK), lactate dehydrogenase (LDH) levels ( W=530.0, P=0.003; W=496.5, P=0.015; W=496.0, P=0.015; W=535.5, P=0.002; W=545.5, P=0.001; W=498.5, P=0.013; W=537.0, P=0.002; W=523.0, P=0.004; W=512.0, P=0.007; W=502.0, P=0.012; W=486.0, P=0.023; W=509.0, P=0.008; W=541.0, P=0.002) and significantly lower platelet count (PLT), indirect bilirubin (IBIL), albumin/globulin ratio(A/G) and superoxide dismutase (SOD) levels ( W=199.0, P=0.024; W=175.5.5, P=0.009; t=-2.9, P=0.004; W=209.5, P=0.036; t=-3.0, P=0.004). ROC result showed that ALP [area under the curve (AUC)=0.804, 95% confidence interval ( CI) (0.679~0.929)] and LDH [AUC=0.805, 95% CI (0.680~ 0.930)] have a higher value for predicting the risk of severe illness. Conclusions:Abnormal liver function, heart function, and renal function indicators in SFTS patients indicate that patients are at risk of exacerbation. Among them, ALP and LDH levels have higher predictive value for risk of severe disease, suggesting that the monitoring of patients with the above symptoms should be strengthened in the clinical nursing process.

Objective:To study the basic characteristics and gene evolution of M segment of fever and thrombocytopenia syndrome virus (SFTSV)in China.Methods:The full sequence of M fragment of SFTSV strain isolated in China before June 25, 2020 was obtained from GenBank. The BioEdit software was used for multi-sequence alignment and MEGA5.0 software was used to construct the phylogenetic tree to compare the homology of SFTSV, which came from different types and sources.Results:M fragments of 203 SFTSV isolated in China was collected from GenBank database.. Among them, 185 strains were isolated from humans (91.1%), 11 strains were isolated from ticks(5.4%), and 5 strains were isolated from goat, mice and hedgehog (3.5%). Phylogenetic tree analysis showed that C2 genotype was dominant in China, accounting for 42.4% of total. The nucleotide sequence and amino acid sequence homology were 96.4%~100.0%, 94.5%~100.0%, respectively. The homology of human and animal isolates was very high. The nucleotide sequence and amino acid sequence homology were 92.5%~100.0% and 92.0%~100.0%, respectively.Conclusions:Phylogenetic tree analysis of SFTSV isolates in China based on the M fragment showed that C2 genotype was dominant, and there are multiple types of co-circulation. By comparing M and S-segment-based phylogenetic tree, it suggests that SFTSV may potential undergo adaptive conversion from C-type to J-type.

Objective To study the role of Med19 in bladder cancer by analyzing the effects of lentivirus-mediated suppression of Med19 expression on T24 bladder cancer cells in vitro.Methods The lentivirus vectors containing a small hairpin RNA (shRNA) to target Med19 were constructed.After T24 bladder cancer cells were infected,real-time PCR and Western-blotting were used to study the Med19 expressions in the CON group (non-infected cells),the NC group (Lv-NC-infected cells) and the KD group (Lv-shMed19-infected cells).The influence of Med19 on the proliferation of bladder cancer cells were assessed using MTT,BrdU,colony formation assay and tumorigenicity experiment in mice.Cell cycle was analyzed with flow cytometry assay.Results Med19 relative mRNA level (0.35 ± 0.03) and Med19 protein expressing in the KD group were significantly inhibited (P ＜ 0.05).The KD group displayed an increased proportion of cells (77.50 ± 0.29)％ in the G0/G1 phase compared with the CON group (69.81 ± 0.81)％and NC group (67.53 ± 0.67) ％ (P ＜ 0.05).Compared with the CON group and the NC group,the KD group displayed a significant cell proliferation defect by MTT and BrdU assay and the number of colonies (91.33 ± 6.11) was significant decreased (P ＜ 0.05).On the day 24,the tumor volume (596.64 ± 485.36) mm3 and weight (0.57 ± 0.44) g of the KD group mice were decreased after inoculation into nude mice (P ＜ 0.05).Specific lentivirus-mediated knockdown of Med19 significantly impacted the cell cycle and proliferation of bladder cancer cells.Infected T24 cells nearly lost their tumorigenicity when being inoculated into nude mice.Conclusion Our results provide new evidence of an important role for Med19 in the development of bladder cancer,suggesting that lentiviruses delivering shRNA against Med19 may be a promising tool for bladder cancer therapy.

Objective To observe the effect of ulinastatin on tumor necrosis factor(TNF-α)and interleukin-6(IL-6)in radiation-induced lung injury.Methods Severity-two female SD rats were randomly divided into 3 groups as control group,irradiation group and treatment group(administered with Ulinastatin).Rats in irradiation group and treatment group were irradiated with linear accelerator at a single dose of 25 Gy.After irradiation rats in treatment group were injected daily with ulinastatin at a dose of 100000 U-kg-1·d-1 for 7 days through caudal vein while rats in control group and irradiation group were injected with the same volume of saline.Rats were killed at 2 h,4,8 and 24 weeks.Samples of lung tissues were observed by using HE staining.Expression of TNF-α in lung was determined by Western blot and expression of IL-6 in serum was determined by ELISA.Data were analyzed by SPSS software.Results Expressions of TNF-α in lung and IL-6 in serum increased significantly after irradiated in irradiation group compared with control group,and it reached the peak at 4 weeks(q=5.63、6.21,P＜0.01).Though expressions of TNF-α and IL-6 in ffeatment group also increased compared with control group,the difference between irradiation group and treatment group was statistic significantly(q=4.97、7.42,P＜0.01).Conclusions TNF-α and IL-6 play an important role in radiation-induced lung injury.Ulinastatin could suppress the inflammatory response and radiation-induced lung injury effectively by decreasing the levels of TNF-α and IL-6.