JMIR Publications

ABSTRACT

Background:

The maintenance and progression of pregnancy rely on the immune homeostasis at the maternal-fetal interface. However, pregnancy complicated with autoimmune abnormalities can disrupt this balance and significantly increase the risk of adverse pregnancy outcomes(APOs).

Objective:

This study aimed to (1) developed an interpretable predictive tool for APOs in patients with immune abnormalities; (2) interpret the models using Shapley Additive Explanations values.

Methods:

This study retrospectively analyzed clinical data from 288 patients with autoimmune abnormalities at Yichang Central People's Hospital between 2019 and 2024. Feature selection was performed using both Boruta algorithm and Lasso regression to identify optimal predictive factors associated with APOs. Nine machine learning (ML) models were developed and subsequently underwent comprehensive comparative evaluation of their predictive performance, leading to identification of the optimal predictive model. Shapley Additive Explanations (SHAP) values were generated to provide interpretable insights into model predictions.

Results:

A total of 288 patients were included in the study, 124 (43.06%) of them had APOs. XGBoost algorithm was shown to be the optimal model after a comparison of nine different models utilizing metrics. The SHAP analysis showed that crown-rump length at 6+0～8+6weeks, number of other drugs, number of complications during pregnancy, gestational sac volume at 6+0～8+6weeks, yolk sac diameter change at 6+0～8+6weeks were the key predictive factors affecting APOs.

Conclusions:

The study developed an interpretable predictive tool for APOs in patients with immune abnormalities, which may assist clinicians in making early intervention decisions.