Retrospective Study
Copyright ©The Author(s) 2021. Published by Baishideng Publishing Group Inc. All rights reserved.
World J Clin Cases. Jun 26, 2021; 9(18): 4573-4584
Published online Jun 26, 2021. doi: 10.12998/wjcc.v9.i18.4573
Application of intelligent algorithms in Down syndrome screening during second trimester pregnancy
Hong-Guo Zhang, Yu-Ting Jiang, Si-Da Dai, Ling Li, Xiao-Nan Hu, Rui-Zhi Liu
Hong-Guo Zhang, Yu-Ting Jiang, Xiao-Nan Hu, Rui-Zhi Liu, Center for Reproductive Medicine and Center for Prenatal Diagnosis, First Hospital, Jilin University, Changchun 130021, Jilin Province, China
Si-Da Dai, Ling Li, College of Communication Engineering, Jilin University, Changchun 130012, Jilin Province, China
Author contributions: Zhang HG and Dai SD contributed to data interpretation and manuscript writing; Jiang YT and Dai SD analyzed the data; Jiang YT and Hu XN contributed to data collection; Liu RZ and Li L contributed to the study design and reviewed the manuscript.
Supported by Science and Technology Department of Jilin Province, No. 20190302073GX.
Institutional review board statement: This study was approved by the Ethics Committee of the First Hospital of Jilin University, No. 2018-387.
Informed consent statement: Informed consent for this study was not requiredas the clinical data were anonymous.
Conflict-of-interest statement: The authors declare that they have no competing interests.
Data sharing statement: No additional data are available.
Open-Access: This article is an open-access article that was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution NonCommercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/Licenses/by-nc/4.0/
Corresponding author: Rui-Zhi Liu, MD, Professor, Center for Reproductive Medicine and Center for Prenatal Diagnosis, First Hospital, Jilin University, No. 1 Xinmin Street, Chaoyang District, Changchun 130021, Jilin Province, China. lrz410@126.com
Received: July 22, 2020
Peer-review started: July 22, 2020
First decision: December 21, 2020
Revised: December 25, 2020
Accepted: March 10, 2021
Article in press: March 10, 2021
Published online: June 26, 2021
ARTICLE HIGHLIGHTS
Research background

Down syndrome (DS) is one of the most common chromosomal aneuploidy diseases. Due to the limitations of DS screening technology, approximately 30% of DS cases could not be found. Therefore, the detection rate and false positive rate of these methods need to be improved.

Research motivation

Issues in recent years have included how to fully utilize clinical cumulative data to provide consultation reference and rational basis for patients and construction of statistical models for DS screening that are suitable for specific regions.

Research objectives

This study aimed to use intelligent algorithms in machine learning for modeling and analysis of prenatal DS screening.

Research methods

This was a retrospective study of a clinical prenatal screening dataset. We designed and developed intelligent algorithms based on the synthetic minority over-sampling technique(SMOTE)-Tomek and adaptive synthetic sampling over-sampling techniques. The machine learning model was established and used for DS screening evaluation.

Research results

The dataset showed a large difference between the numbers of DS affected and non-affected cases. A combination of over-sampling and under-sampling techniques can greatly increase the performance of the algorithm at processing non-balanced datasets. As the number of iterations increases, the combination of the classification and regression tree algorithm and the SMOTE-Tomek over-sampling technique can obtain a high detection rate (DR) while keeping the false positive rate(FPR) to a minimum.

Research conclusions

Intelligent algorithms achieved good results on the DS screening dataset. When the T21 risk cutoff value was set to 270, machine learning methods had a higher DR and a lower FPR than statistical methods.

Research perspectives

The findings of this study suggest that the establishment and application of machine learning models will help to improve the detection rate of DS.