Improved support vector machine algorithm based on the influence of Gestational Diabetes Mellitus on the outcome of perinatal outcome by ultrasound imaging

基于妊娠期糖尿病对围产期结局影响的超声成像改进支持向量机算法

阅读:1

Abstract

OBJECTIVES: In order to understand the incidence and epidemiological characteristics of gestational diabetes mellitus, the ultrasound imaging of support vector machine processing algorithm was used to clarify the outcome of maternal and neonatal gestational diabetes mellitus. METHODS: This study selected clinical data of 12,190 pregnant women who were hospitalized for delivery, and were divided into diabetic group (1268 cases) and control group (10922 cases) according to the diagnosis of gestational diabetes. The study was conducted from January 1, 2012 to December 31, 2019. Colour Doppler ultrasound was performed to record fatal umbilical artery and brain the middle arteries and uterine arteries which are effective indicators of measuring fatal intrauterine conditions. Chi-square test was used to compare the rates between groups, and multivariate logistic regression was used for labour outcomes. RESULTS: The incidence of diabetes during pregnancy is about 10.4% (1268/12190). Senior citizens and women suffering from obesity increase the risk of gestational diabetes, maternal hypertension disorders in pregnancy, premature rupture of membranes, oligohydramnios, fatal distress, multiple births, malpresentation risk increased significantly (P <0.05) than the control group. In gestational diabetes caesarean section rate was significantly higher (61.0% vs46.4%). Caesarean new born 5-minute Apgar score was significantly lower than the control group (P <0.05). CONCLUSION: In maternal gestational diabetes in high risk pregnancies, complications of pregnancy significantly increase the importance of enhancing weight management and blood glucose monitoring to reduce complications.

特别声明

1、本页面内容包含部分的内容是基于公开信息的合理引用;引用内容仅为补充信息,不代表本站立场。

2、若认为本页面引用内容涉及侵权,请及时与本站联系,我们将第一时间处理。

3、其他媒体/个人如需使用本页面原创内容,需注明“来源:[生知库]”并获得授权;使用引用内容的,需自行联系原作者获得许可。

4、投稿及合作请联系:info@biocloudy.com。