Predicting Spontaneous Pre-term Birth Risk Is Improved When Quantitative Ultrasound Data Are Included With Historical...

Area of Science:

• Obstetrics and Gynecology
• Medical Imaging
• Biostatistics

Background:

• Predicting spontaneous pre-term birth (sPTB) is challenging due to a lack of early warning signs.
• Current methods relying on historical clinical (HC) data have limited predictive accuracy.

Purpose of the Study:

• To evaluate if combining HC data with quantitative ultrasound (QUS) data improves the prediction of sPTB risk.
• To compare the predictive performance of a model using HC data alone versus a model using combined HC and QUS data.

Main Methods:

• A study involving 248 full-term births and 26 sPTBs.
• Quantitative ultrasound (QUS) cervical scans were performed at 20 ± 2 and 24 ± 2 weeks of gestation.
• Two statistical models were compared: HC data alone and combined HC and QUS data, using receiver operating characteristic area under the curve (AUC) for evaluation.

Main Results:

• The combined HC and QUS data model achieved a significantly better AUC (0.68) for sPTB prediction compared to HC data alone (0.53).
• The inclusion of QUS features in the classification model was statistically significant (p < 0.01).
• Even with a small number of sPTBs, QUS data added value to risk prediction.

Conclusions:

• Combining quantitative ultrasound data with historical clinical data enhances the prediction of spontaneous pre-term birth risk.
• This integrated approach offers a more effective tool for identifying women at risk for sPTB.