A novel MRI-based diagnostic model for predicting placenta accreta spectrum.
Jianfeng Xia1, Yongren Hu1, Zehe Huang1
1Department of Radiology, The First People's Hospital of Qinzhou, 53500, China.
Magnetic Resonance Imaging
|February 26, 2024
Summary
A new diagnostic model using MRI signs effectively predicts placenta accreta spectrum (PAS). This model, combining clinical factors and MRI features, shows strong predictive performance in both training and validation sets.
Area of Science:
- Radiology
- Obstetrics & Gynecology
- Medical Imaging
Background:
- Placenta accreta spectrum (PAS) is a significant obstetric complication.
- Accurate prediction of PAS is crucial for effective management and improved maternal outcomes.
- Magnetic Resonance Imaging (MRI) offers detailed anatomical visualization for diagnosing pelvic conditions.
Purpose of the Study:
- To develop and evaluate a diagnostic model for predicting placenta accreta spectrum (PAS).
- To assess the predictive performance of clinical factors, MRI signs, and a combined model for PAS.
Main Methods:
- A study involving 155 pregnant women, with 62 diagnosed with PAS and 93 without.
- Development of diagnostic models using logistic regression with clinical factors and/or MRI features.
- Evaluation of model performance using Receiver Operating Characteristic (ROC) curve analysis and validation sets.
Main Results:
- Four clinical factors and 11 MRI features were identified as significant risk factors for PAS.
- The combined clinical and MRI model demonstrated the highest diagnostic performance (AUC 0.874 in training, 0.735 in validation).
- MRI-based models showed superior predictive accuracy compared to clinical factors alone.
Conclusions:
- Diagnostic models incorporating MRI features can effectively predict placenta accreta spectrum.
- The combined clinical and MRI model offers the most robust prediction of PAS.
- MRI is a valuable tool in the preoperative assessment of suspected placenta accreta spectrum.
