Advances in Artificial Intelligence Models for Diagnosis and Prediction Based on MRI in Placenta Accreta Spectrum
10.3969/j.issn.1005-5185.2025.11.018
- VernacularTitle:基于MRI的胎盘植入性疾病人工智能诊断与预测模型研究进展
- Author:
Aitong LI
1
;
Tao LU
Author Information
1. 电子科技大学医学院,四川 成都 610054
- Publication Type:Journal Article
- Keywords:
Placenta accreta spectrum;
Magnetic resonance imaging;
Artificial intelligence;
Radiomics;
Deep learning;
Diagnosis;
Forecasting
- From:
Chinese Journal of Medical Imaging
2025;33(11):1241-1244
- CountryChina
- Language:Chinese
-
Abstract:
Placenta accreta spectrum,characterized by abnormal placental adhesion or invasion into the uterine myometrium,represent severe obstetric complications that may lead to uterine rupture and life-threatening postpartum hemorrhage.Although ultrasound remains the primary screening modality for placenta accreta spectrum,its diagnostic accuracy is limited by operator experience and gestational age.MRI with its advantages of multi-sequence acquisition,multiplanar capabilities,high soft-tissue resolution and wide field of view,serves as a critical adjunct for diagnosis when ultrasound results are inconclusive.However,MRI interpretation heavily relies on radiologists'expertise.The integration of artificial intelligence(AI)with MRI has emerged as a promising approach to address this challenge.This article reviews recent advances in MRI-based AI diagnostic and predictive models for placenta accreta spectrum,focusing on the application of radiomics and deep learning techniques in AI-driven diagnostic and predictive models.The technical strengths,clinical applicability and limitations of these models are critically analyzed,and future directions for enhancing generalizability and clinical translation are proposed.
