OBJECTIVE:Nowadays,machine learning algorithms are gradually being applied to predict stroke and cardiovascular disease.Compared with traditional regression models,machine learning can learn from data to achieve high prediction accuracy by exploring the flexible relationship between a large number of predictive features and outcome variables,providing a new method for the formulation of individualized treatment and rehabilitation programs.This study aims to systematically evaluate stroke functional recovery and prognosis prediction models based on machine learning,comprehensively assessing their predictive performance and clinical application potential to provide references for the development,application,and promotion of related predictive models.METHODS:This review was conducted following the PRISMA(Preferred Reporting Items for Systematic Reviews and Meta-Analyses)guidelines.Relevant literature on stroke prognosis prediction using machine learning methods was selected by searching PubMed,EMbase,Web of Science Core Collection,CNKI,WanFang,and the China Biomedical Literature Database,with the search period from January 1,2014,to July 1,2024.Two researchers independently screened the literature and extracted data based on inclusion and exclusion criteria,using the Prediction model Risk Of Bias ASsessment Tool(PROBAST)to assess model quality.RESULTS:(1)A total of 3 126 articles were obtained in the preliminary search.After screening and exclusion,18 articles were finally included.150 prediction models were constructed using 13 machine learning methods.The three most frequently used methods are Logistic Regression,Random Forest,and Extreme Gradient Boosting(XGBoost).Only one study was externally validated.Eight studies reported how the missing data were handled.(2)In terms of outcome indicators,8 studies used the combination of clinical data and imaging data to build models,9 studies only used clinical data to build models,and 1 study only used imaging data to build models.(3)Each of the 18 studies gave the most important characteristics of the study,with the most mentioned being the National Institute of Health Stroke Scale and age.All studies reported area under curve values ranging from 0.74 to 0.96,with the highest area under curve being 0.96.The overall risk of bias in all models was high.The high risk of bias in the field of model analysis was the main reason for the high risk of overall bias in all models.(4)The results of meta-analysis showed that age and National Institute of Health Stroke Scale score had significant influence on stroke prognosis,with age[MD=8.49,95％CI(6.24,10.75),P＜0.01]and National Institute of Health Stroke Scale score[MD=4.78,95％CI(2.56,7.00),P＜0.01].CONCLUSION:This study systematically evaluated the predictive model of functional recovery and prognosis of stroke based on machine learning,and all the models have good predictive potential.However,future studies should increase the sample size of the included model,adopt prospective studies,and add external validation of the model to improve the stability and prediction accuracy of the model,control the risk of bias,and contribute to the validation and promotion of the model in practical clinical applications.At the same time,the interpolation of missing values is more transparent and accurate.Although existing machine learning models show good predictive performance,it is also important to focus on the functionality and usability of the model,and the inclusion of features will reduce ease of use.We should develop easy to use model interfaces and user-friendly clinical tools to enable medical staff to better apply the model for clinical decision.

OBJECTIVE:Nowadays,machine learning algorithms are gradually being applied to predict stroke and cardiovascular disease.Compared with traditional regression models,machine learning can learn from data to achieve high prediction accuracy by exploring the flexible relationship between a large number of predictive features and outcome variables,providing a new method for the formulation of individualized treatment and rehabilitation programs.This study aims to systematically evaluate stroke functional recovery and prognosis prediction models based on machine learning,comprehensively assessing their predictive performance and clinical application potential to provide references for the development,application,and promotion of related predictive models.METHODS:This review was conducted following the PRISMA(Preferred Reporting Items for Systematic Reviews and Meta-Analyses)guidelines.Relevant literature on stroke prognosis prediction using machine learning methods was selected by searching PubMed,EMbase,Web of Science Core Collection,CNKI,WanFang,and the China Biomedical Literature Database,with the search period from January 1,2014,to July 1,2024.Two researchers independently screened the literature and extracted data based on inclusion and exclusion criteria,using the Prediction model Risk Of Bias ASsessment Tool(PROBAST)to assess model quality.RESULTS:(1)A total of 3 126 articles were obtained in the preliminary search.After screening and exclusion,18 articles were finally included.150 prediction models were constructed using 13 machine learning methods.The three most frequently used methods are Logistic Regression,Random Forest,and Extreme Gradient Boosting(XGBoost).Only one study was externally validated.Eight studies reported how the missing data were handled.(2)In terms of outcome indicators,8 studies used the combination of clinical data and imaging data to build models,9 studies only used clinical data to build models,and 1 study only used imaging data to build models.(3)Each of the 18 studies gave the most important characteristics of the study,with the most mentioned being the National Institute of Health Stroke Scale and age.All studies reported area under curve values ranging from 0.74 to 0.96,with the highest area under curve being 0.96.The overall risk of bias in all models was high.The high risk of bias in the field of model analysis was the main reason for the high risk of overall bias in all models.(4)The results of meta-analysis showed that age and National Institute of Health Stroke Scale score had significant influence on stroke prognosis,with age[MD=8.49,95％CI(6.24,10.75),P＜0.01]and National Institute of Health Stroke Scale score[MD=4.78,95％CI(2.56,7.00),P＜0.01].CONCLUSION:This study systematically evaluated the predictive model of functional recovery and prognosis of stroke based on machine learning,and all the models have good predictive potential.However,future studies should increase the sample size of the included model,adopt prospective studies,and add external validation of the model to improve the stability and prediction accuracy of the model,control the risk of bias,and contribute to the validation and promotion of the model in practical clinical applications.At the same time,the interpolation of missing values is more transparent and accurate.Although existing machine learning models show good predictive performance,it is also important to focus on the functionality and usability of the model,and the inclusion of features will reduce ease of use.We should develop easy to use model interfaces and user-friendly clinical tools to enable medical staff to better apply the model for clinical decision.

AIM: To observe the clinical and multimodal imaging features of retinopathy associated with novel coronavirus disease 2019(COVID-19)infection, investigate the related risk factors, and analyze the treatment and prognosis.METHODS: A total of 7 patients(7 eyes)with clinically confirmed COVID-19-associated retinopathy in Tianjin Medical University General Hospital from December 13, 2022 to January 13, 2023 were included in the study. All patients underwent color fundus photography, IR, spectral-domain optical coherence tomography(SD-OCT), fundus autofluorescein(FAF)and other ophthalmic examination and serological examination.RESULTS: Among the included patients, 2 cases(2 eyes)of central retinal vein occlusion(CRVO)appeared differently from previous CRVO. The hemorrhagic features of CRVO were round or fusiform shape hemorrhagic spots with white centers. One of them, the von Willebrand factor antigen(vWF: Ag)level was increased to 161.8%. The other case was positive in serologic test for lupus anticoagulant. In 2 cases(2 eyes)of multiple evanescent white dot syndrome(MEWDS), FAF showed that dots of high spontaneous fluorescence were scattered in the posterior pole. The prognosis of 2 MEWDS were good after the treatment of glucocorticoids. The 3 cases(3 eyes)of acute macular neuroretinopathy(AMN)showed reddened brown lesions in the macular area, hyporeflective lesions with clear boundaries on IR, and high signal intensity in the ONL and ELM, EZ/IZ signal fracture on SD-OCT.CONCLUSION: COVID-19 may cause inflammatory storm, involving all layers of retinal tissues and blood vessels, leading to the occurrence of various retinal lesions. Hormone therapy may be effective and the prognosis is good in most cases. Roth spot can be seen in fundus hemorrhage of CRVO, lupus anticoagulant and increased vWF: Ag may be risk factors for CRVO after COVID-19.

Objective:To explore the value of high-resolution CT (HRCT) visual scores and quantitative analysis in assessing pulmonary Langerhans cell histiocytosis (PLCH) in adults.Methods:In total 51 adult patients with PLCH confirmed by pathology in Peking Union Medical College Hospital from August 2014 to December 2021 were retrospectively analyzed. All patients underwent HRCT and pulmonary function tests (PFT). The involvement of the nodular and cystic lesions were evaluated by two experienced radiologists using CT visual scores. The cases were divided into three groups based on the nodular scores, and into four groups based on the cystic scores, respectively. Ratio of low attenuation areas (LAA%) was measured by an automatic post-processing software. Pulmonary function indices including forced expiratory volume in the first second (FEV 1), forced vital capacity (FVC), FEV 1/FVC, diffusion capacity for carbon monoxide of lung (D LCO), alveolar ventilation (V A), D LCO/V A, D LCO corrected for hemoglobin (D LCOc), D LCOc/V A were collected. FEV 1/FVC was expressed as measured values and other indices were expressed as percent predicted (%pred). Spearman correlation analysis was used to evaluate the correlation between HRCT visual scores, LAA% and PFT. The lung function indices among different nodular groups as well as among different cystic groups were compared using the Kruskal‐Wallis test. Results:Both nodular and cystic lesions were found on HRCT images of all 51 patients. There were no correlation between the visual scores of nodular lesions and lung function indices (all P>0.05). There were no significant differences in lung function indices among different nodular groups (all P>0.05). The visual scores of cystic lesions were negatively correlated with FEV 1/FVC, D LCO%pred, D LCO/V A%pred, D LCOc%pred, D LCOc/V A%pred ( r=-0.491, -0.347, -0.330, -0.373, -0.346, respectively, all P<0.05); the pulmonary function indices among different cystic groups had significant difference (all P<0.05). LAA% were negatively correlated with FEV 1/FVC, D LCO%pred, D LCO/V A%pred, D LCOc%pred, D LCOc/V A%pred ( r=-0.278, -0.378, -0.418, -0.395, -0.451, respectively, all P<0.05). Conclusion:HRCT visual scores of nodular lesions do not correlate with lung function in patients with PLCH. Visual scores and quantitative analysis of the cystic lesions can reflect the impairment degree of pulmonary ventilation and diffusion function to a certain extent, and may be used in assessment of patients with PLCH.

Objective To investigate the value of hepatocyte growth factor ( HGF ) and transforming growth factor-β1 (TGF-β1) in the diagnosis of early stage non-small cell lung cancer (ES-NSCLC) through detection of their levels in serum and bronchoalveolar lavage fluid ( BALF ) . Methods Serum and BALF samples were collected from 48 cases of ES-NSCLC ( at stage 1 and 2 ) , 45 cases of medium and advanced NSCLC (at stage 3 and 4), 42 cases of benign pulmonary nodule and 30 controls for the study. The levels of HGF and TGF-β1 were respectively detected by ELISA, and their differences were analyzed between the 4 groups. The relationship between HGF and TGF-β1 levels and tumor size, differentiation and pathological type was further analyzed, and sensitivity and specificity in the diagnosis of ES-NSCLC was evaluated in the study. Results The levels of HGF and TGF-β1 in BALF of the ES-NSCLC group were significantly higher than those in the benign pulmonary nodule group and the normal control group (P <0. 05). HGF and TGF-β1 serum levels in the ES-NSCLC group were higher than those in the benign pulmonary nodule group and the control group, however, no statistical significance could be seen when comparisons were made between them ( P >0. 05). Serum HGF and TGF-β1 levels in the BALF of the medium and advanced NSCLC group were significantly higher than those in the ES-NSCLC group (P<0. 05). The areas under the curve (AUC) in the BALF of HGF and TGF-1 were respectively 0. 754 and 0. 839. When the levels of 59. 32pg/ml and 98. 52pg/ml were selected as the optimal cutoff values according to the Jordan index maximum method, the sensitivity of HGF and TGF-β1 in the diagnosis of ES-NSCLC were respectively 41. 7％ and 81. 3％, and specificity were respectively 96. 3％ and 77. 8％. Furthermore, in the patients with adenocarcinoma or tumor size larger than 3cm in diameter, higher expression levels could be detected. Conclusion High expression levels of HGF and TGF-β1 could be detected in the BALF of the patients with ES-NSCLC with high specificity. For this reason, it has certain clinical value in the accessory diagnosis of ES-NSCLC.

Objective To investigate the value of hepatocyte growth factor ( HGF ) and transforming growth factor-β1 (TGF-β1) in the diagnosis of early stage non-small cell lung cancer (ES-NSCLC) through detection of their levels in serum and bronchoalveolar lavage fluid ( BALF ) . Methods Serum and BALF samples were collected from 48 cases of ES-NSCLC ( at stage 1 and 2 ) , 45 cases of medium and advanced NSCLC (at stage 3 and 4), 42 cases of benign pulmonary nodule and 30 controls for the study. The levels of HGF and TGF-β1 were respectively detected by ELISA, and their differences were analyzed between the 4 groups. The relationship between HGF and TGF-β1 levels and tumor size, differentiation and pathological type was further analyzed, and sensitivity and specificity in the diagnosis of ES-NSCLC was evaluated in the study. Results The levels of HGF and TGF-β1 in BALF of the ES-NSCLC group were significantly higher than those in the benign pulmonary nodule group and the normal control group (P <0. 05). HGF and TGF-β1 serum levels in the ES-NSCLC group were higher than those in the benign pulmonary nodule group and the control group, however, no statistical significance could be seen when comparisons were made between them ( P >0. 05). Serum HGF and TGF-β1 levels in the BALF of the medium and advanced NSCLC group were significantly higher than those in the ES-NSCLC group (P<0. 05). The areas under the curve (AUC) in the BALF of HGF and TGF-1 were respectively 0. 754 and 0. 839. When the levels of 59. 32pg/ml and 98. 52pg/ml were selected as the optimal cutoff values according to the Jordan index maximum method, the sensitivity of HGF and TGF-β1 in the diagnosis of ES-NSCLC were respectively 41. 7％ and 81. 3％, and specificity were respectively 96. 3％ and 77. 8％. Furthermore, in the patients with adenocarcinoma or tumor size larger than 3cm in diameter, higher expression levels could be detected. Conclusion High expression levels of HGF and TGF-β1 could be detected in the BALF of the patients with ES-NSCLC with high specificity. For this reason, it has certain clinical value in the accessory diagnosis of ES-NSCLC.

OBJECTIVETo use a hospital-based case-control study design to investigate the relationship between hepatocellular carcinoma (HCC) and the interaction of polymorphisms in the human mismatch repair gene,hMSH2,with environmental factors.

METHODSCases of new-onset,histopathology-diagnosed,and untreated (no chemotherapy or radiation therapy) HCC were enrolled between September 2009 and September 2012.A non-HCC healthy control group was also enrolled and was composed of individuals living in the same region as the cases for more than 10 years and age-/sex-matched with similar socioeconomic characteristics.All enrollees underwent hMSH2 genotyping by real-time PCR.T-test,chi-square test and unconditional logistic regression analysis was used to analyze the difference in allele frequencies among the case and control groups and the relationship between hMSH2 polymorphisms and environmental factors.

RESULTSFrequencies of hMSH2 rs2303428 CC,CT and TT genotypes in the HCC group were significantly different than in the control group (14.13% vs.8.21%,47.02% vs.49.47%,and 38.85% vs.42.32%;x 2=8.289,P =0.016).Individuals carrying the hMSH2 rs2303428 T allelic gene had a significantly increased risk compared to those with the hMSH2 rs2303428 C allelic gene (adjusted OR=1.228).Interactions were found between the hMSH2 genotype and hepatitis B surface antigen (HBsAg)-positive hepatitis infection (adjusted OR=1.865) and history of cancer (adjusted OR=5.634).There was no relation between hMSH2 gene rs4952887 and rs2059520 and liver cancer development or interaction with environmental factors.

CONCLUSIONThe hMSH2 rs2303428 genotype is positively related to risk of HCC in Chinese,with HBsAg-positive hepatitis infection starus and history of cancer increasing the risk.

Alleles ; Carcinoma, Hepatocellular ; genetics ; Case-Control Studies ; Female ; Gene Frequency ; Genotype ; Humans ; Liver Neoplasms ; genetics ; Male ; MutS Homolog 2 Protein ; genetics ; Polymorphism, Genetic ; Real-Time Polymerase Chain Reaction