The Combined Use of Cervical Ultrasound and Deep Learning Improves the Detection of Patients at Risk for Spontaneous Preterm Delivery

BACKGROUND: Preterm birth is the leading cause of neonatal mortality and morbidity. While ultrasound-based cervical length measurement is the current standard for predicting preterm birth, its performance is limited. Artificial intelligence has shown potential in ultrasound analysis, yet few small-scale studies have evaluated its use for predicting preterm birth.

OBJECTIVE: To develop and validate an artificial intelligence model for spontaneous preterm birth prediction from cervical ultrasound images and compare its performance to cervical length.

STUDY DESIGN: In this multicenter study, we developed a deep learning-based artificial intelligence model using data from women who underwent cervical ultrasound scans as part of antenatal care between 2008 and 2018 in Denmark. Indications for ultrasound were not systematically recorded, and scans were likely performed due to risk factors or symptoms of preterm labor. We compared the performance of the artificial intelligence model with cervical length measurement for spontaneous preterm birth prediction by assessing the area under the curve, sensitivity, specificity, and likelihood ratios. Subgroup analyses evaluated model performance across baseline characteristics, and saliency heat maps identified anatomical features that influenced artificial intelligence model predictions the most.

RESULTS: The final dataset included 4224 pregnancies and 7862 cervical ultrasound images, with 50% resulting in spontaneous preterm birth. The artificial intelligence model surpassed cervical length for predicting spontaneous preterm birth before 37 weeks with a sensitivity of 0.51 (95% confidence interval, 0.50-0.53) versus 0.41 (0.39-0.42) at a fixed specificity at 0.85, P<0.001 and a higher area under the curve of 0.75 (0.74-0.76) versus 0.67 (0.66-0.68), P<0.001. For identifying late preterm births at 34 to 37 weeks, the artificial intelligence model had 36.6% higher sensitivity than cervical length (0.47 versus 0.34, P<0.001). The artificial intelligence model achieved higher area under the curves across all subgroups, especially at earlier gestational ages. Saliency heat maps indicated that in 70% of preterm birth cases, the artificial intelligence model focused on the inner lining of the lower uterine segment, suggesting it incorporates more data than cervical length alone.

CONCLUSION: To our knowledge, this is the first large-scale multicenter study demonstrating that artificial intelligence is more sensitive than cervical length measurement in identifying spontaneous preterm births across multiple characteristics, 19 hospital sites, and different ultrasound machines. The artificial intelligence model performs particularly well at earlier gestational ages, enabling more timely prophylactic interventions.