IAS-FET: An intelligent assistant system and an online platform for enhancing successful rate of in-vitro fertilization embryo transfer technology based on clinical features

Ying Ma; Bowen Zhang; Zhaoqing Liu; Yujie Liu; Jiarui Wang; Xingxuan Li; Fan Feng; Yali Ni; Shuyan Li

doi:10.1016/j.cmpb.2024.108050

IAS-FET: An intelligent assistant system and an online platform for enhancing successful rate of in-vitro fertilization embryo transfer technology based on clinical features

Comput Methods Programs Biomed. 2024 Mar:245:108050. doi: 10.1016/j.cmpb.2024.108050. Epub 2024 Jan 24.

Authors

Ying Ma¹, Bowen Zhang², Zhaoqing Liu³, Yujie Liu³, Jiarui Wang³, Xingxuan Li⁴, Fan Feng¹, Yali Ni¹, Shuyan Li⁵

Affiliations

¹ Gansu Provincial Maternity and Child-care Hospital, Lanzhou, Gansu 730030, China.
² School of Medical Information and Engineering, Xuzhou Medical University, Xuzhou, Jiangsu 221004, China; School of Electronic Information and Communications, Huazhong University of Science and Technology, Wuhan, Hubei 430073, China.
³ School of Medical Information and Engineering, Xuzhou Medical University, Xuzhou, Jiangsu 221004, China.
⁴ School of Chemistry and Chemical Engineering, Lanzhou University, Lanzhou, Gansu 730030, China.
⁵ School of Medical Information and Engineering, Xuzhou Medical University, Xuzhou, Jiangsu 221004, China. Electronic address: lishuyan@xzhmu.edu.cn.

PMID: 38301430
DOI: 10.1016/j.cmpb.2024.108050

Abstract

Background: Among all of the assisted reproductive technology (ART) methods, in vitro fertilization-embryo transfer (IVF-ET) holds a prominent position as a key solution for overcoming infertility. However, its success rate hovers at a modest 30% to 70%. Adding to the challenge is the absence of effective models and clinical tools capable of predicting the outcome of IVF-ET before embryo formation. Our study is dedicated to filling this critical gap by aiming to predict IVF-ET outcomes and ultimately enhance the success rate of this transformative procedure.

Methods: In this retrospective study, infertile patients who received artificial assisted pregnancy treatment at Gansu Provincial Maternity and Child-care Hospital in China were enrolled from 2016 to 2020. Individual's clinical information were studied by cascade XGBoost method to build an intelligent assisted system for predicting the outcome of IVF-ET, called IAS-FET. The cascade XGBoost model was trained using clinical information from 2292 couples and externally tested using clinical information from 573 couples. In addition, several schemes which will be of help for patients to adjust their physical condition to improve their success rate on ART were suggested by IAS-FET.

Results: The outcome of IVF-ET can be predicted by the built IAS-FET method with the area under curve (AUC) value of 0.8759 on the external test set. Besides, this IAS-FET method can provide several schemes to improve the successful rate of IVF-ET outcomes. The built tool for IAS-FET is addressed as a free platform online at http://www.cppdd.cn/ART for the convenient usage of users.

Conclusions: It suggested the significant influence of personal clinical features for the success of ART. The proposed system IAS-FET based on the top 27 factors could be a promising tool to predict the outcome of ART and propose a plan for the patient's physical adjustment. With the help of IAS-FET, patients can take informed steps towards increasing their chances of a successful outcome on their journey to parenthood.

Keywords: Clinical features; In vitro fertilization-embryo transfer technology; Machine learning; Smart and interactive platform.

MeSH terms

Embryo Transfer* / methods
Female
Fertilization
Fertilization in Vitro* / methods
Humans
Pregnancy
Pregnancy Rate
Retrospective Studies