Automated Approach for Enhancing Fetal Head Station Assessment in Labor with Transperineal Ultrasound

Area of Science:

• Obstetrics and Gynecology
• Medical Imaging
• Computational Medicine

Background:

• Accurate fetal head station (FHS) assessment is critical for labor management and delivery planning.
• Digital vaginal examination (DVE) for FHS assessment has known inaccuracies.
• Ultrasound (US) evaluation of FHS is operator-dependent, relying on sonographer expertise.

Purpose of the Study:

• To evaluate the reliability and accuracy of an automated algorithm for FHS assessment using transperineal ultrasound (TPU).
• To compare the performance of the automated FHS assessment with conventional DVE and expert sonographer assessment.

Main Methods:

• A prospective observational study involving 27 pregnant women in the second stage of labor.
• Transperineal ultrasound (TPU) acquisitions were performed, and FHS was assessed by an automated algorithm (FHS_aut) and digital vaginal examination (FHS_dig).
• Expert sonographer assessment (FHS_exp) on TPU images served as the gold standard; performance was analyzed using correlation coefficients, R², RMSE, and Bland-Altman analysis.

Main Results:

• The automated algorithm showed a strong correlation with expert assessment (ρ = 0.97, R² = 0.95).
• Digital vaginal examination demonstrated a weaker correlation with expert assessment (ρ = 0.66, R² = 0.52).
• The automated method achieved higher accuracy with lower RMSE (0.32 cm) compared to DVE (0.97 cm), with smaller bias and narrower limits of agreement.

Conclusions:

• The proposed automated algorithm accurately assesses fetal head station during labor with high reliability.
• This automated TPU approach has the potential to enhance clinical decision-making and support labor management.
• The technology could facilitate broader adoption of ultrasound in labor monitoring, improving patient care.