BACKGROUND: Over 20 million colonoscopies are performed in China annually. An automatic clinical decision support system (CDSS) with accurate semantic recognition of colonoscopy reports and guideline-based is helpful to relieve the increasing medical burden and standardize the healthcare. In this study, the CDSS was built under a hierarchical-label interpretable classification framework, trained by a state-of-the-art transformer-based model, and validated in a multi-center style. METHODS: We conducted stratified sampling on a previously established dataset containing 302,965 electronic colonoscopy reports with pathology, identified 2041 patients' records representative of overall features, and randomly divided into the training and testing sets (7:3). A total of five main labels and 22 sublabels were applied to annotate each record on a network platform, and the data were trained respectively by three pre-training models on Chinese corpus website, including bidirectional encoder representations from transformers (BERT)-base-Chinese (BC), the BERT-wwm-ext-Chinese (BWEC), and ernie-3.0-base-zh (E3BZ). The performance of trained models was subsequently compared with a randomly initialized model, and the preferred model was selected. Model fine-tuning was applied to further enhance the capacity. The system was validated in five other hospitals with 3177 consecutive colonoscopy cases. RESULTS: The E3BZ pre-trained model exhibited the best performance, with a 90.18% accuracy and a 69.14% Macro-F1 score overall. The model achieved 100% accuracy in identifying cancer cases and 99.16% for normal cases. In external validation, the model exhibited favorable consistency and good performance among five hospitals. CONCLUSIONS The novel CDSS possesses high-level semantic recognition of colonoscopy reports, provides appropriate recommendations, and holds the potential to be a powerful tool for physicians and patients. The hierarchical multi-label strategy and pre-training method should be amendable to manage more medical text in the future.
Humans ; Colonoscopy/methods* ; Deep Learning ; Decision Support Systems, Clinical ; Female ; Male

OBJECTIVE To construct and evaluate nomogram prediction model for refractory chemotherapy-induced nausea and vomiting （CINV）. METHODS The data of malignant tumor patients who received chemotherapy at the Third People’s Hospital of Zhengzhou from January 2017 to December 2023 were collected. These patients were categorized into the occurrence group and the non-occurrence group according to the occurrence of refractory CINV. Multivariate Logistic regression analysis was employed to screen predictive factors for refractory CINV and constructing a nomogram prediction model. Model performance was assessed via receiver operating characteristic curve analysis. Model calibration was evaluated using Bootstrap resampling. Decision curve analysis （DCA） was used to determine the clinical net benefit of three strategies under different risk thresholds. Clinical impact curves were utilized to assess the clinical value of the model at different risk thresholds. Shapley additive explanations （SHAP） analysis was performed to evaluate individual factor contributions to the predictive model. RESULTS A total of 388 patients were included， with 219 experiencing refractory CINV. Multivariate Logistic regression identified 11 predictive factors for refractory CINV， including gastrointestinal disease history， anticipated nausea and vomiting， chemotherapy-induced emetic risk classification， and electrolyte levels， etc. The model’s area under the curve was 0.80 [95% confidence interval （0.76， 0.84）]， with a mean error of 0.036. DCA demonstrated the prediction model had higher clinical net benefit when the risk threshold was between 0.05 and 0.85. SHAP analysis revealed the top three predictive factors as gastrointestinal disease history （0.924）， chemotherapy- induced emetic risk classification （0.866）， and electrolyte levels （0.581）. CONCLUSIONS Eleven factors， including gastrointestinal disease history， anticipated nausea and vomiting， chemotherapy-induced emetic risk classification， and electrolyte levels， are identified as predictors of refractory CINV. The model based on these factors has good predictive ability， which can be used to predict the risk of refractory CINV.

Objective To investigate the effect and mechanism of melatonin in alleviating hypoxia-induced apoptosis in ovarian gra-nulosa cells.Methods Rat ovarian granulosa cells were isolated and divided into normoxic,hypoxic,and melatonin groups.Hypoxia-induced injury models were established in the hypoxic and melatonin groups,and granulosa cells in the melatonin group were treated with melatonin.A total of 24 rats were randomized into the control,model,and intervention groups(n=8 per group).Rat models of declining ovarian function induced by long-term hypoxia were established in the model and intervention groups.The rats in the intervention group were intraperitoneally injected with melatonin.Cell proliferation was measured using a CCK-8 assay,and lactate secretion and HIF-1αprotein with a specific kit,respectively.The levels of estradiol and progesterone in the cell supernatant and rat serum were detected using ELISA.Granulosa cell apoptosis was detected by flow cytometry,ovarian morphology by HE staining,and Bax and caspase-3 expression by Western blotting.Results Compared with the normoxic group,the hypoxic group exhibited decreased granulosa cell proliferation,increased apoptosis,elevated lactate and HIF-1α levels,and reduced estradiol and progesterone levels(P＜0.05).Compared with hypoxic group,these changes were significantly reversed in the molatonin group(P＜0.05).Compared with the control group,the model group showed increased lactate,HIF-1α,Bax,and caspase-3 levels,decreased estradiol and progesterone levels,and reduced follicles.Compared with the model group,all the indicators were ameliorated in the intervention group(P＜0.05).Conclusion Melatonin alleviated hypoxia-induced granulosa cell apoptosis and promoted the recovery of ovarian function.

Objective To investigate the effect and mechanism of melatonin in alleviating hypoxia-induced apoptosis in ovarian gra-nulosa cells.Methods Rat ovarian granulosa cells were isolated and divided into normoxic,hypoxic,and melatonin groups.Hypoxia-induced injury models were established in the hypoxic and melatonin groups,and granulosa cells in the melatonin group were treated with melatonin.A total of 24 rats were randomized into the control,model,and intervention groups(n=8 per group).Rat models of declining ovarian function induced by long-term hypoxia were established in the model and intervention groups.The rats in the intervention group were intraperitoneally injected with melatonin.Cell proliferation was measured using a CCK-8 assay,and lactate secretion and HIF-1αprotein with a specific kit,respectively.The levels of estradiol and progesterone in the cell supernatant and rat serum were detected using ELISA.Granulosa cell apoptosis was detected by flow cytometry,ovarian morphology by HE staining,and Bax and caspase-3 expression by Western blotting.Results Compared with the normoxic group,the hypoxic group exhibited decreased granulosa cell proliferation,increased apoptosis,elevated lactate and HIF-1α levels,and reduced estradiol and progesterone levels(P＜0.05).Compared with hypoxic group,these changes were significantly reversed in the molatonin group(P＜0.05).Compared with the control group,the model group showed increased lactate,HIF-1α,Bax,and caspase-3 levels,decreased estradiol and progesterone levels,and reduced follicles.Compared with the model group,all the indicators were ameliorated in the intervention group(P＜0.05).Conclusion Melatonin alleviated hypoxia-induced granulosa cell apoptosis and promoted the recovery of ovarian function.

Objective:To investigate the value of systemic inflammatory response syndrome (SIRS), pediatric sequential organ failure assessment (pSOFA) and pediatric critical illness score (PCIS) in predicting mortality of pediatric sepsis in pediatric intensive care units (PICU) from Southwest China.Methods:This was a prospective multicenter observational study. A total of 447 children with sepsis admitted to 12 PICU in Southwest China from April 2022 to March 2023 were enrolled. Based on the prognosis, the patients were divided into survival group and non-survival group. The physiological parameters of SIRS, pSOFA and PCIS were recorded and scored within 24 h after PICU admission. The general clinical data and some laboratory results were recorded. The area under the curve (AUC) of the receiver operating characteristic curve was used to compare the predictive value of SIRS, pSOFA and PCIS in mortality of pediatric sepsis.Results:Amongst 447 children with sepsis, 260 patients were male and 187 patients were female, aged 2.5 (0.8, 7.0) years, 405 patients were in the survival group and 42 patients were in the non-survival group. 418 patients (93.5%) met the criteria of SIRS, and 440 patients (98.4%) met the criteria of pSOFA≥2. There was no significant difference in the number of items meeting the SIRS criteria between the survival group and the non-survival group (3(2, 4) vs. 3(3, 4) points, Z=1.30, P=0.192). The pSOFA score of the non-survival group was significantly higher than that of the survival group (9(6, 12) vs. 4(3, 7) points, Z=6.56, P<0.001), and the PCIS score was significantly lower than that of the survival group (72(68, 81) vs. 82(76, 88) points, Z=5.90, P<0.001). The predictive value of pSOFA (AUC=0.82) and PCIS (AUC=0.78) for sepsis mortality was significantly higher than that of SIRS (AUC=0.56) ( Z=6.59, 4.23, both P<0.001). There was no significant difference between pSOFA and PCIS ( Z=1.35, P=0.176). Platelet count, procalcitonin, lactic acid, albumin, creatinine, total bilirubin, activated partial thromboplastin time, prothrombin time and international normalized ratio were all able to predict mortality of sepsis to a certain degree (AUC=0.64, 0.68, 0.80, 0.64, 0.68, 0.60, 0.77, 0.75, 0.76, all P<0.05). Conclusion:Compared with SIRS, both pSOFA and PCIS had better predictive value in the mortality of pediatric sepsis in PICU.

Objective: To understand sexual orientation and characteristics of sexual behaviors related with HIV infection among HIV positive male college students in Beijing and to provide suggestions for development of sexual health education and HIV prevention strategy for student population. Methods: HIV positive male college students diagnosed from 2016 to June 30, 2019 in Beijing were recruited. Questionnaire was used to retrospectively investigate sexual orientation and characteristics of sexual behaviors before HIV infection. Behaviors related to HIV infection were compared between absolutely homosexual and others sexual orientation. Results: The average age of male students firstly identified to be HIV positive was (22.18±2.70) years old, 61.5%(123) of them were undergraduate, 69.5%(139) were not registered permanent residency in Beijing, 38.5%(77) of their interviewed sexual partners were from online chat and 83.0%(166) had homosexual behaviors. Sexual orientation score analysis showed that 50.0% of the participants self identified as exclusively homosexual. Compared with other sexual oriental group, exclusively homosexual group had lower mean age of their HIV infection firstly identified（ t =2.77， P =0.01）, higher rate of Rush use, firstly insertive sexual behavior with male, firstly insertive sexual behavior before 18 years old, sexual partners more than three persons, having regular partners, nonpersistent use of condom, being diagnosed of sexual transmitted disease and the frequency of homosexual behaviors more than 1 time per month （ χ 2=5.15,28.06,4.16,5.34,5.89,7.39,6.68, P <0.05）. Rush users had higher rate of STD diagnosis than non users in exclusively homosexual group ( χ 2=6.26， P =0.01). Conclusion Risky sexual behaviors associated with HIV infection were higher in exclusively homosexual group then other sexual orientation groups among HIV positive male college students. Family and school should concern with sexual health education byreinforcing health education via network media to improve college students’ awarenees on HIV/AIDS.

OBJECTIVE To explore the influence of surgical history on chemotherapy -induced nausea and vomiting （CINV）. METHODS A retrospective case -control study was adopted ，with 824 patients undergoing chemotherapy as the object . A total of 27 items were collected ，including demographic data ，medical history data ，pre-chemotherapy data ，and chemotherapy treatment status. Logistic regression model was used to analyze the relationship between the history of surgery and the risk of CINV . The multiple models were constructed to correct potential confounding factors ，and subgroup analysis was performed on patients with surgical history . RESULTS The incidence of CINV was higher in patients with surgical history . The statistical result before adjustment was [OR＝1.72，95%CI（1.31，2.28），P＜0.001]；after adjusting potential confounding factors ，the statistical result was [OR＝1.78，95% CI（1.28，2.48），P＝0.001]. In the subgroup analysis ，the time between surgery and chemotherapy was different ， and the impact of surgical history on CINV was different ，and the results were statistically significant （P＝0.027）. The risk of CINV showed decreasing trend with the time ，and the results were statistically significant （P for trend ≤0.050）. Compared with patients who had not undergone surgery ，patients who had undergone surgery within one year had a higher risk of CINV [OR＝ 2.33，95%CI（1.52，3.59），P＜0.001]. CONCLUSIONS Patients with surgical history are more prone to CINV ，and the risk of CINV shows a downward trend in the length of time from surgery .

Objective::To evaluate the agreement and reliability of intrapartum nonreasurring cardiotocography (CTG) interpretation and prediction of neonatal acidemia by obstetricians working in different centers.Methods::A retrospective cohort study involving two tertiary hospitals (The First Affiliated Hospital of Sun Yat-sen University and Perking University Third Hospital) was conducted between 30 th September 2018 and 1 st April 2019. Six obstetricians from two hospitals with three levels of experience (junior, medium, and senior) reviewed 100 nonreassuring fetal heart rate (FHR) tracings from 1 hour before the onset of abnormalities until delivery. Each reviewer determined the FHR pattern, the baseline, variability, and presence of acceleration, deceleration, sinusoidal pattern, and predicted whether neonatal acidemia and abnormal umbilical arterial pH < 7.1 would occur. Inter-observer agreement was assessed using the proportions of agreement (Pa) and the proportion of specific agreement (Pa for each category). Reliability was evaluated with the kappa statistic (k-Light’s kappa for n raters) and Gwet’s AC1 statistic. Results::Good inter-observer agreement was found in evaluation of most variables (Pa>0.5), with the exception of early deceleration (Pa= 0.39, 95% confidence interval ( CI): 0.36,0.43). Reliability was also good among most variables (AC1 > 0.40), except for acceleration, early deceleration, and prediction of neonatal acidemia (AC1 = 0.17, 0.10, and 0.25, respectively). There were no statistically significant differences among the three groups, except in the identification of accelerations (Pa = 0.89, 95% CI: 0.83,0.95; Pa= 0.50, 95% CI: 0.41,0.60, and Pa= 0.35, 95% CI: 0.25,0.43 in junior, medium and senior groups, respectively) and the prediction of neonatal acidemia (Pa= 0.52, 0.52, and 0.62 in junior, medium and senior groups, respectively), where agreement was highest and lowest in the junior-level group, respectively. The accuracy and sensitivity of the prediction for umbilical artery pH < 7.1 were similar among the three groups, but the specificity was higher in the senior groups (93.68% vs. 92.53% vs. 98.85% in junior, medium and senior groups, P= 0.015). Conclusion::Although we found a good inter-observer agreement in the evaluation of the most basic CTG features and FHR category statistically, it was insufficient to meet the clinical requirements for "no objection" interpretation for FHR tracings. Further specialized training is needed for standardized interpretation of intrapartum FHR tracings.

Objective::To evaluate the agreement and reliability of intrapartum nonreasurring cardiotocography (CTG) interpretation and prediction of neonatal acidemia by obstetricians working in different centers.Methods::A retrospective cohort study involving two tertiary hospitals (The First Affiliated Hospital of Sun Yat-sen University and Perking University Third Hospital) was conducted between 30 th September 2018 and 1 st April 2019. Six obstetricians from two hospitals with three levels of experience (junior, medium, and senior) reviewed 100 nonreassuring fetal heart rate (FHR) tracings from 1 hour before the onset of abnormalities until delivery. Each reviewer determined the FHR pattern, the baseline, variability, and presence of acceleration, deceleration, sinusoidal pattern, and predicted whether neonatal acidemia and abnormal umbilical arterial pH < 7.1 would occur. Inter-observer agreement was assessed using the proportions of agreement (Pa) and the proportion of specific agreement (Pa for each category). Reliability was evaluated with the kappa statistic (k-Light’s kappa for n raters) and Gwet’s AC1 statistic. Results::Good inter-observer agreement was found in evaluation of most variables (Pa>0.5), with the exception of early deceleration (Pa= 0.39, 95% confidence interval ( CI): 0.36,0.43). Reliability was also good among most variables (AC1 > 0.40), except for acceleration, early deceleration, and prediction of neonatal acidemia (AC1 = 0.17, 0.10, and 0.25, respectively). There were no statistically significant differences among the three groups, except in the identification of accelerations (Pa = 0.89, 95% CI: 0.83,0.95; Pa= 0.50, 95% CI: 0.41,0.60, and Pa= 0.35, 95% CI: 0.25,0.43 in junior, medium and senior groups, respectively) and the prediction of neonatal acidemia (Pa= 0.52, 0.52, and 0.62 in junior, medium and senior groups, respectively), where agreement was highest and lowest in the junior-level group, respectively. The accuracy and sensitivity of the prediction for umbilical artery pH < 7.1 were similar among the three groups, but the specificity was higher in the senior groups (93.68% vs. 92.53% vs. 98.85% in junior, medium and senior groups, P= 0.015). Conclusion::Although we found a good inter-observer agreement in the evaluation of the most basic CTG features and FHR category statistically, it was insufficient to meet the clinical requirements for "no objection" interpretation for FHR tracings. Further specialized training is needed for standardized interpretation of intrapartum FHR tracings.

OBJECTIVE: To describe the clinical and genetic characteristics of a child with 14q12q13.1 deletion involving the FOXG1 gene. METHODS: Clinical manifestation of the child was analyzed. Peripheral blood sample of the patient was subjected to chromosomal karyotyping and single nucleotide polymorphism array (SNP-array) analysis. RESULTS: The male infant has developed feeding difficulty, poor sucking, lower limb tremor, and frontal bruising 8 days after birth. Magnetic resonance imaging revealed significant enlargement of bilateral ventricles and corpus callosum dysplasia. Chromosomal analysis revealed a karyotype of 46,XY,del(14)(q12q13.1), and SNP-array confirmed that there was a 9.6 Mb deletion in 14q11.2q13.1, which encompassed the FOXG1 gene. CONCLUSION For patients with brain development abnormalities, dyskinesia, cognitive impairment, speech disorder and other manifestations, copy number variation of the FOXG1 gene should be excluded. SNP-array should be carried out as early as possible to attain the diagnosis.
Child ; Chromosome Deletion ; DNA Copy Number Variations ; Forkhead Transcription Factors/genetics* ; Heterozygote ; Humans ; Infant ; Karyotyping ; Male ; Nerve Tissue Proteins/genetics* ; Polymorphism, Single Nucleotide