Objective:To evaluate the accuracy of the oral healthcare information provided by different large language models (LLM) to explore their feasibility and limitations in the application of oral auxiliary, treatment and health consultation.Methods:This study designed eight items comprising 47 questions in total related to the diagnosis and treatment of oral diseases [to assess the performance of LLM as an artificial intelligence (AI) medical assistant], and five items comprising 35 questions in total about oral health consultations (to assess the performance of LLM as a simulated doctor). These questions were answered individually by the five LLM models (Erine Bot, HuatuoGPT, Tongyi Qianwen, iFlytek Spark, ChatGPT). Two attending physicians with more than 5 years of experience independently rated the responses using the 3C criteria (correct, clear, concise), and the consistency between the raters was assessed using the Spearman rank correlation coefficient, and the Kruskal-Wallis test and Dunn post hoc test were used to assess the statistical differences between the models. Additionally, this study used 600 questions from the 2023 dental licensing examination to evaluate the time taken to answer, scores, and accuracy of each model.Results:As an AI medical assistant, LLM can assist doctors in diagnosis and treatment decision-making, with an inter-evaluator Spearman coefficient of 0.505 ( P<0.01). As a simulated doctor, LLM can carry out patient popularization, with an inter-evaluator Spearman coefficient of 0.533 ( P<0.01). The 3C scores of each model as an AI medical assistant and a simulated doctor were respectively: 2.00 (1.00, 3.00) and 2.00 (2.00, 3.00) points of Erine Bot, 1.00 (1.00, 2.00) and 2.00 (1.00, 2.00) points of HuatuoGPT, 2.00 (1.00, 2.00) and 2.00 (1.00, 3.00) points of Tongyi Qianwen, 2.00 (1.00, 2.00) and 2.00 (1.75, 2.25) points of iFlytek Spark, 3.00 (2.00, 3.00) and 3.00 (2.00, 3.00) points of ChatGPT (full score of 4 points). The Kruskal-Wallis test results showed that, as an AI medical assistant or a simulated doctor, there were statistically differences in the 3C scores among the five large language models (all P<0.001). The average score of the 5 LLMs on the dental licensing examination was 370.2, with an accuracy rate of 61.7% (370.2/600) and a time consumption of 94.6 min. Specifically, Erine Bot took 115 min, scored 363 points with an accuracy rate of 60.5% (363/600), HuatuoGPT took 224 min and scored 305 points with an accuracy rate of 50.8% (305/600), Tongyi Qianwen took 43 min, scored 438 points with an accuracy rate of 73.0% (438/600), iFlytek Spark took 32 min, scored 364 points with an accuracy rate of 60.7% (364/600), and ChatGPT took 59 min, scored 381 points with an accuracy rate of 63.5% (381/600). Conclusions:Based on the evaluation of LLM′s dual roles as an AI medical assistant and a simulated doctor, ChatGPT performes the best, with basically correct, clear and concise answers, followed by Erine Bot, Tongyi Qianwen and iFlytek Spark, with HuatuoGPT lagging behind significantly. In the dental licensing examination, all the 4 LLM, except for HuatuoGPT, reach the passing level, and the time consumpution for answering is significantly reduced compared to the 8 h required for the exam regulations in all of the five models. LLM has the feasibility of application in oral auxiliary, treatment and health consultation, and it can help both doctors and patients obtain medical information quickly. Howere, their outputs carry a risk of errors (since the 3C scoring results do not reach the full marks), so prudent judgment should be exercised when using them.

Objective:To evaluate the accuracy of the oral healthcare information provided by different large language models (LLM) to explore their feasibility and limitations in the application of oral auxiliary, treatment and health consultation.Methods:This study designed eight items comprising 47 questions in total related to the diagnosis and treatment of oral diseases [to assess the performance of LLM as an artificial intelligence (AI) medical assistant], and five items comprising 35 questions in total about oral health consultations (to assess the performance of LLM as a simulated doctor). These questions were answered individually by the five LLM models (Erine Bot, HuatuoGPT, Tongyi Qianwen, iFlytek Spark, ChatGPT). Two attending physicians with more than 5 years of experience independently rated the responses using the 3C criteria (correct, clear, concise), and the consistency between the raters was assessed using the Spearman rank correlation coefficient, and the Kruskal-Wallis test and Dunn post hoc test were used to assess the statistical differences between the models. Additionally, this study used 600 questions from the 2023 dental licensing examination to evaluate the time taken to answer, scores, and accuracy of each model.Results:As an AI medical assistant, LLM can assist doctors in diagnosis and treatment decision-making, with an inter-evaluator Spearman coefficient of 0.505 ( P<0.01). As a simulated doctor, LLM can carry out patient popularization, with an inter-evaluator Spearman coefficient of 0.533 ( P<0.01). The 3C scores of each model as an AI medical assistant and a simulated doctor were respectively: 2.00 (1.00, 3.00) and 2.00 (2.00, 3.00) points of Erine Bot, 1.00 (1.00, 2.00) and 2.00 (1.00, 2.00) points of HuatuoGPT, 2.00 (1.00, 2.00) and 2.00 (1.00, 3.00) points of Tongyi Qianwen, 2.00 (1.00, 2.00) and 2.00 (1.75, 2.25) points of iFlytek Spark, 3.00 (2.00, 3.00) and 3.00 (2.00, 3.00) points of ChatGPT (full score of 4 points). The Kruskal-Wallis test results showed that, as an AI medical assistant or a simulated doctor, there were statistically differences in the 3C scores among the five large language models (all P<0.001). The average score of the 5 LLMs on the dental licensing examination was 370.2, with an accuracy rate of 61.7% (370.2/600) and a time consumption of 94.6 min. Specifically, Erine Bot took 115 min, scored 363 points with an accuracy rate of 60.5% (363/600), HuatuoGPT took 224 min and scored 305 points with an accuracy rate of 50.8% (305/600), Tongyi Qianwen took 43 min, scored 438 points with an accuracy rate of 73.0% (438/600), iFlytek Spark took 32 min, scored 364 points with an accuracy rate of 60.7% (364/600), and ChatGPT took 59 min, scored 381 points with an accuracy rate of 63.5% (381/600). Conclusions:Based on the evaluation of LLM′s dual roles as an AI medical assistant and a simulated doctor, ChatGPT performes the best, with basically correct, clear and concise answers, followed by Erine Bot, Tongyi Qianwen and iFlytek Spark, with HuatuoGPT lagging behind significantly. In the dental licensing examination, all the 4 LLM, except for HuatuoGPT, reach the passing level, and the time consumpution for answering is significantly reduced compared to the 8 h required for the exam regulations in all of the five models. LLM has the feasibility of application in oral auxiliary, treatment and health consultation, and it can help both doctors and patients obtain medical information quickly. Howere, their outputs carry a risk of errors (since the 3C scoring results do not reach the full marks), so prudent judgment should be exercised when using them.

Objective To investigate the effects of tegafur, gimeracil and oteracil potassium combined with oxaliplatin on gastric motility-related hormones, matrix metalloproteinase-2 (MMP-2) and matrix metalloproteinase-9 (MMP-9) in elderly patients with gastric cancer. Methods A total of 128 elderly patients with gastric cancer were selected as the study subjects and randomly divided into control group (n=64) and observation group (n=64). The control group was treated with tegafur, gimeracil and oteracil potassium, while the observation group received tegafur, gimeracil and oteracil potassium combined with oxaliplatin. Gastric motility-related hormones [motilin (MTL) and vasoactive intestinal peptide (VIP)], MMP-2, MMP-9, tumor markers [carcinoembryonic antigen (CEA), carbohydrate antigen 125 (CA125) and carbohydrate antigen 19-9 (CA19-9)], clinical efficacy and occurrence of adverse reactions were observed in both groups. The correlations of gastric motility-related hormones with the levels of MMP-2, MMP-9 and tumor markers were analyzed. Results After treatment, MTL levels in both groups were significantly lower, while VIP levels were significantly higher compared to before treatment. Additionally, the observation group had significantly lower MTL and higher VIP levels compared to the control group (P < 0.05). After treatment, the levels of MMP-2, MMP-9, CEA, CA125 and CA19-9 were significantly reduced in both groups compared to pretreatment levels, with the observation group showing significantly lower levels than the control group (P < 0.05). The total effective rate in the observation group was 70.31%, which was significantly higher than the 53.13% in the control group (P < 0.05). There was no statistically significant difference in the incidence of adverse reactions between the two groups (P>0.05). A negative correlation was observed between MTL and MMP-2, MMP-9, CEA, CA125 as well as CA19-9 (P < 0.05), while VIP showed a positive correlation with MMP-2, MMP-9, CEA, CA125 and CA19-9 (P < 0.05). Conclusion The combination of tegafur, gimeracil and oteracil potassium and oxaliplatin is effective and safe for the treatment of elderly patients with gastric cancer. MTL, VIP, MMP-2, MMP-9, CEA, CA125 and CA19-9 are interrelated in the development of gastric cancer in elderly patients.

Objective:To evaluate the effect of esketamine combined with ultrasound-guided dorsal penile nerve block (DPNB) on negative postoperative behavioral changes (NPOBCs) in pediatric patients undergoing circumcision under general anesthesia.Methods:One-hundred and ninety-five pediatric patients, aged 4-8 yr, with body mass index of 10-35 kg, of American Society of Anesthesiologists Physical Status classificationⅠ or Ⅱ, undergoing elective circumcision under general anesthesia, were selected and divided into 3 groups ( n=65 each) using a random number table method: esketamine group (group E), DPNB group (group D) and esketamine combined with DPNB group (group ED). Propofol 1.5 mg/kg was intravenously injected, and the patients were admitted to the operating room after consciousness disappeared in the 3 groups. Esketamine 0.5 mg/kg was intravenously injected in E and ED groups, and the equal volume of normal saline was given in group D. D and ED groups underwent bilateral DPNB with 0.25 % ropivacaine 0.15 ml/kg under ultrasound guidance, with the maximum total amount of the drug not exceeding 10 ml. Fentanyl 1.0 μg/kg and propofol 2.0 mg/kg were intravenously injected prior to the skin incision in the three groups. If intraoperative body movement occurred, propofol 10 mg was added, which could be repeated. The occurrence of intraoperative body movement, respiratory depression and amount of propofol added was recorded. When postoperative pain (FLACC score >4) occurred, flurbiprofen 1 mg/kg was intravenously injected for analgesia, and the usage of flurbiprofen was recorded. When emergence agitation(PEAD score>10) occurred, propofol 1 mg/kg was intravenously injected for sedation, and the occurrence of emergence agitation was recorded. Parents were followed up by telephone at 1, 7 and 30 days postoperatively to assess the occurrence of NPOBCs using the PHBQ scale. Results:Fifty-six patients in group E and 59 patients in D and ED groups finally completed the study.Compared with group E, the incidence of intraoperative body movement was significantly decreased, the amount of additional propofol was reduced, the emergence agitation score, incidence of emergence agitation and severe agitation and usage rate of postoperative flurbiprofen were decreased, and the incidence of separation anxiety at 7 and 30 days postoperatively was decreased in D and ED groups, and the incidence of intraoperative respiratory depression was significantly decreased, and the incidence of NPOBCs at 7 and 30 days postoperatively was decreased in group ED ( P<0.05). Compared with group D, the incidence of intraoperative respiratory depression was significantly decreased, the amount of additional propofol was decreased, the usage rate of postoperative flurbiprofen and incidence of sleep anxiety at 1 day postoperatively were decreased ( P<0.05), and no significant change was found in the incidence of NPOBCs at each time point after operation in group ED ( P>0.05). Conclusions:Esketamine combined with ultrasound-guided DPNB can reduce the occurrence of NPOBCs in pediatric patients undergoing circumcision under general anesthesia.

Objective:To investigate the effect of acute hypervolemic hemodilution (AHH) with 6% hydroxyethyl starch 130/0.4 on pharmacodynamics of propofol during successful laryngeal mask airway (LMA) implantation.Methods:American Society of Anesthesiology physical status Ⅰ or Ⅱ patients, aged 30-60 yr, with body mass index of 18.5-25.0 kg/m 2, undergoing elective extensive total hysterectomy under general anesthesia, were divided into 2 groups: AHH group (group A) and control group (group C). In group A, 6% hydroxyethyl starch 130/0.4 was infused at a rate of 20 ml/min for AHH, and the target hematocrit was 30%.In group C, lactated Ringer′s solution was infused according to the " 4-2-1" rule to supplement physiological requirements, and anesthesia induction was performed after 10 min of stabilization.Sufentanil was administered by target-controlled infusion using Bovil pharmacokinetic model with effect-site concentration (Ce) of 0.25 ng/ml, 3 min later propofol was given by target-controlled infusion using Schnider model.The Ce of propofol in the first patient was set at 5.0 μg/ml.Each time the concentration of propofol was increased/decreased by 0.5 μg/ml according to the sequential method.LMA was inserted following 1 min equilibration between plasma concentration and Ce of propofol.The trial was terminated when 8 consecutive inflection points of failed/successful LMA insertion occurred.The EC 5, EC 50, EC 95 and 95% confidence interval (95% CI) of propofol were calculated by probit regression analysis. Results:In group A, the EC 5 (95% CI), EC 50 (95% CI) and EC 95 (95% CI) of propofol when LMA was successfully placed were 4.237 (3.090-4.514) μg/ml, 4.802 (4.500-5.078) μg/ml and 5.443 (5.125-7.304) μg/ml, respectively.In group C, the EC 5 (95% CI), EC 50 (95% CI) and EC 95 (95% CI) of propofol when LMA was successfully placed were 2.408 (1.190-2.756) μg/ml, 3.120 (2.690-3.472) μg/ml and 4.042 (3.582-7.431) μg/ml, respectively.There was significant difference in EC 5, EC 50 and EC 95 between the two groups ( P<0.01). Conclusion:AHH with 6% hydroxyethyl starch 130/0.4 can decrease the efficacy of propofol when LMA is successfully implanted.

Objective To investigate the distribution and characteristics of iodine excess areas in Liaocheng City of Shandong Province, and to provide data evidence for taking intervention measures. Methods From 2011 to 2013, 1 - 3 samples of drinking water were collected from all administrative villages in 8 counties (cities and districts) of Liaocheng.At the same time,1 sample of edible salt was collected from the household where water samples were collected. Arsenic and cerium spectrophotometry was used for the detection of water iodine and salt iodine was detected by semi-quantitative method. The region were divide according to the definition of "Water Source Excess Iodine Area and Excess Iodine Disease Area"(GB/T 19380-2016)and"Division of Iodine Deficiency Disorders Area"(GB 16005-2009).Results A total of 7 794 water samples were collected in 5 865 villages of 134 towns and the iodine median was 158.2 μg/L. The median of water iodine of 57 samples was less than 10 μg/L in drinking water and the ratio was 0.7%;2 286 samples were 10- 100 μg/L and the ratio was 29.3%; 5 451 samples were over 100 μg/L and the ratio was 69.9%. The towns with suitable water iodine (10 - 100 μg/L) and high water iodine ( > 100 μg/L) were 24.6%(33/134)and 75.4%(101/134), respectively, and no iodine deficiency town was found. The areas with high water iodine were distributed in patchy or foci. A total of 3 300 salt samples were collected,among them,iodized salt was 1 183(35.58%,1 183/3 300)and non-iodized salt was 2 117(64.15%,2 117/3 300). Among them, there were 36 towns with high iodine content and 20 towns with suitable iodine content, and the iodized salt coverage rates were 10.72% (225/2 099) and 79.77% (958/1 201),respectively.Conclusions The population of Liaocheng City is at risk of iodine excess.The high iodine areas coexists with suitable iodine areas.

ABSTRACT:Objective To investigate the interaction of polymorphisms of TNF-αgene promoter-308G/A and PPAR-γ2 gene-C34G with acute pancreatitis (AP)and its severity degree.Methods Totally 150 mild acute pancreatitis(MAP),150 moderately severe acute pancreatitis(MSAP)and 150 severe acute pancreatitis(SAP)cases were selected for this study,and 450 healthy persons as control group.The genetic polymorphisms of TNF-αgene promoter-308G/A and PPAR-γ2 gene-C34G were analyzed by the technique of PCR in peripheral blood leukocytes of above-mentioned cases and the results were verified by direct DNA sequencing method.Results The frequencies of -308G/A(GA),-308G/A(AA),-C34G(CG)and-C34G(GG)were 24.00%,26.67%,24.00% and 26.00% in MAP group,34.67%,36.67%,34.00% and 36.67% in MSAP group,42.00%,46.00%,43.33% and 46.00% in SAP group,and 14.44%,14.22%,12.89% and 14.67% in control group,respectively.Statistical tests showed significant difference in the frequencies among each group (all P<0 .0 1 ).The risk of AP significantly increased in subjects with-308G/A(GA),genotype (ORMAP=2.677 6,ORMSAP=6.625 0,ORSAP=21.514 7),in those with-308G/A(AA)genotype (ORMAP=2.570 0,ORMSAP=6.401 8,ORSAP=18.903 4),in those with-C34G(CG) genotype (ORMAP=2.668 4,ORMSAP=6.776 9,ORSAP=22.207 2),and in those with-C34G(GG)genotype (ORMAP=2.633 8,ORMSAP=6.472 5,ORSAP=21.570 2).Combined analysis of the polymorphisms showed that percentage of-308G/A(AA)/-C34G(GG)in MAP,MSAP,SAP and control groups was 7.33%,13.33%,20.67% and 2.00%,respectively,and statistical tests showed significant difference in the frequency among each group (all P<0.01).The people who carried-308G/A(AA)/-C34G(GG)had a high risk of AP (ORMAP=7.284 2,ORMSAP=41.296 1,ORSAP=363.973 6),and statistical analysis suggested a positive interaction between-308G/A(AA)and-C34G(GG)in increasing the risk of AP (γ2MAP=2.114 2,γ4MAP=2.080 0,γ2MSAP=2.108 7,γ4MSAP=2.050 6,γ2SAP=2.138 8,γ4SAP=2.000 1).Likewise,there were also positive interactions in the pathogenesis of AP between-308G/A(GA)and-C34G(GG),-308G/A(GA)and-C34G(CG),-308G/A(AA)and-C34G(CG)(All γ>1). Conclusion These carriers of-308G/A(GA),-308G/A(AA),-C34G(CG)and-C34G(GG)genotypes may have a high risk of developing AP,and significant interactions between genetic polymorphisms of-308G/A and-C34G add the risk of the occurrence and development of AP.

Objective To evaluate the skill and outcome of arthroscopic reconstruction of posterior cruciate ligament (PCL) with retention of PCL remnant. Methods From April 2004 to June 2006, 38 patients (38 knees) with PCL deficiency were verified by clinical and arthroscopic examinations. Of them, there were 9 knees combined with disruption of the posterolateral comer, 6 with rupture of the posteromedial corner, 8 with lateral meniscus tear and 4 with medial meniscus tear. With reservation of PCL remnant and synovium, all the impaired PCLs were reconstructed with single bundle of autogenous quadrupled hamstring tendons under arthroscopy. Interference screws were used for direct anatomic fixation of the reconstructed ligament. Results No severe comphcations occurred at early stage after operation in all 38 patients who were followed up for 12-37 months (average 20.79 months). Lysholm score was improved significandy from 40-70 points (mean 51.32 pints) before operation to 70-100 pints (mean 92.37 points) at the latest follow up (t=-30.14, P<0.01). According to International Knee Documentation Committee (IKDC) score, there was a remarkable improvement from 16 abnormal knees (grade C) and 22 severely abnormal knees (grade D) preoperatively to 18 normal knees ( grade A), 18 nearly normal knees (grade B) and 2 abnormal knees at the latest follow up (Z=-6.00, P <0.01). Of 38 patients, 36 returned to normal sports level but 2 degraded level of sports. Conclusions Arthroscopic PCL reconstruction with retention of PCL remnants is a feasible technique, with satisfactory outcome. Preservation of PCL remnants and synovium may be beneficial to biological incorporation and reinnervation of the reconstructed ligament.

Objective To observe the effects of estrogen on behavior and 5-HT in periaqueductal gray (PAG)in migraine model rats. Methods Tewnty-four ovariectomized Wistar rats were divided randomly into four groups:control group (Group A),migraine group(Group B),low dose estradiol-treated ovariectomized group(Group C),high dose etradiol-treated ovariectomized group(Group D).After 1 week,the rats in Group B,C and D were injected with nitroglycerine 10 mg?kg-1 subcutaneously to make migraine rat models,the rats in Group A were given peanut oil alike,and the behavior changes were observed.2 h after injection,the rats were killed and the midbrains were separated and then 5-HT immunohistochemical staining was performed.Results Behavior: compared with Group B,the degrees of red-calws,red-ears and red-tail rats in Group D relieved obviously,the times of climbing hutch and scratching head were much fewer,while the rats in group C showed no significant difference;Immunohistochemical staining:compared with Group A,the 5-HT-positive neurons expression in PAG of Group B and C were more obviously(P

Blood Gas Analysis ; Hepatopulmonary Syndrome ; therapy ; Humans ; Male ; Middle Aged ; Portasystemic Shunt, Transjugular Intrahepatic ; Treatment Outcome