Correlation Between Ultrasound Soft Markers And Fetal Chromosome Abnormality

Objective:This study consists of two parts.The first part uses CiteSpace software to visually analyze the hot spots and trends in the research field of ultrasound soft markers and chromosome abnormalities at home and abroad,which provides a theoretical basis for this study.In the second part,the Logistic regression model and CHAID decision tree model were established with 10 common ultrasound soft markers as independent variables and fetal chromosome abnormality as dependent variables.The purpose of this paper is to analyze the influencing factors and effects of fetal chromosome abnormality from different angles,so as to improve the accuracy of screening and avoid unnecessary invasive prenatal diagnosis.Methods:1.Literature analysis: Searching the literature published in CNKI and Web of Science database for the correlation analysis between ultrasound soft markers and chromosome abnormality of super infants.The time is set from January 1,2010 to December 31,2022,and the included literatures are screened according to strict inclusion and elimination criteria.CNKI database uses Refworks format to export documents for data processing,Web of Science database uses plain text file format to export documents for data processing.After using CiteSpace6.1.R6 software to remove duplication and manually screen irrelevant literatures,the amount of published literatures and keywords were analyzed,and the keyword co-occurrence map,cluster map and time zone map were drawn.The hot spots and trends in the research field of ultrasound soft markers and fetal chromosome abnormality at home and abroad in recent years were analyzed.2.Model study: From January 2017 to December 2021,the clinic information of 642 pregnant women in Suining Central Hospital were retrospectively analyzed.All pregnant women had chromosome examination results and were positive for single soft markers by ultrasound examination.By using SPSS 26.0 statistical software,taking the fetal chromosome results as dependent variables,and taking 10 ultrasound soft markers as independent variables,Logistic regression model and the Chi-squared Automatic Interaction Detector(CHAID)decision tree model were established.The ROC curves of Logistic regression model and decision tree model were drawn,and the area under ROC curves of the two prediction models were compared by Z test.The difference was statistically significant with P < 0.05.Results :1.By analyzing 920 literatures in CNKI database,it can be seen that there are many papers published during the four years from 2018 to 2021,and there are many studies on ultrasound soft markers and fetal chromosome abnormalities during this period;Through the keyword map,we can see that the research hotspots and trends in this field mainly focus on:(1)Choose different chromosome diagnosis and screening methods according to different ultrasound soft markers.(2)NT thickening,absent or hypoplastic nasal bone and ventriculomegaly are the most studied ultrasound soft markers.(3)Study on the correlation between ultrasound soft markers and adverse pregnancy outcome.2.By analyzing 323 documents in the Web of Science database,it can be seen that the number of published documents showed a slow upward trend from 2010 to 2022,and the number of published documents was more during the three years from 2020 to 2022;According to the keyword map,the research hotspots in this field mainly focus on:(1)The relationship between positive ultrasound soft markers and fetal chromosome abnormality in early pregnancy,and the combination of ultrasound soft markers and serology in screening Down syndrome in early pregnancy.(2)Research on chromosome detection technology.(3)Ultrasound soft markers are the NT thickening,absent or hypoplastic nasal bone,ventriculomegaly,fetal echogenic bowel and so on.(4)Fetal ultrasound soft markers positive and heart malformation and other aspects of the study.3.(1)Univariate Logistic regression analysis showed that NT thickening,nasal bone absent and ventriculomegaly are the risk factors of fetal chromosome abnormality;Multivariate Logistic regression analysis screened NT thickening(OR=7.511,P < 0.05),nasal bone absent(OR=4.819,P < 0.05)and ventriculomegaly(OR=4.789,P < 0.05)to fit the regression model,The regression equation was obtained:Logit(P)=-2.888+2.016×NT thickening+1.572×nasal bone absent+1.566×ventriculomegaly.(2)CHAID decision tree model showed that NT thickening and nasal bone absent are the influencing factors of fetal chromosome abnormality.(3)The area under ROC curve of Logistic regression model is larger than that of CHAID decision tree model(0.712 vs 0.675,Z=2.267,P < 0.05).Conclusion:1.Searching and analyzing the domestic and foreign literatures,we find that:(1)In recent years,the research heat in this field in China has declined,while the research heat in this field in foreign countries has maintained a continuous upward trend.(2)The study of ultrasound soft markers with high predictive value and the selection of non-invasive means to screen fetal chromosomal abnormalities and structural abnormalities are the current research hotspots.(3)The screening of fetal chromosomal abnormalities in China concentrates mainly in the second trimester,while foreign research trends are more inclined to use ultrasound soft markers combined with serological results of pregnant women to screen Down’s syndrome in early pregnancy,and gradually replace chromosome examination in second pregnancy,so as to detect fetuses with abnormal chromosomes earlier,thus relieving physical pain and mental stress of pregnant women.(4)NT thickening is the most studied ultrasound soft marker at home and abroad.2.Through the analysis of two predictive models,we can draw the following conclusions:(1)NT thickening is of higher value as a predictor of fetal chromosome abnormality,followed by nasal bone loss.(2)The correlation between ultrasound soft markers and fetal chromosome abnormality can be analyzed by logistic regression model and decision tree model.(3)Logistic regression model is superior to decision tree model in the study of positive single ultrasound soft markers and fetal chromosome abnormality.Therefore,the results of this study suggest that a correct understanding of USM with suitable statistical analysis method can provide new ideas for the study of fetal chromosome abnormalities and clinical selection of diagnostic schemes.