Multi-centre deep learning for placenta segmentation in obstetric ultrasound with multi-observer and cross-country generalization

Abstract

The placenta is crucial to fetal well-being and it plays a significant role in the pathogenesis of hypertensive pregnancy disorders. Moreover, a timely diagnosis of placenta previa may save lives. Ultrasound is the primary imaging modality in pregnancy, but high-quality imaging depends on the access to equipment and staff, which is not possible in all settings. Convolutional neural networks may help standardize the acquisition of images for fetal diagnostics. Our aim was to develop a deep learning based model for classification and segmentation of the placenta in ultrasound images. We trained a model based on manual annotations of 7,500 ultrasound images to identify and segment the placenta. The model's performance was compared to annotations made by 25 clinicians (experts, trainees, midwives). The overall image classification accuracy was 81%. The average intersection over union score (IoU) reached 0.78. The model’s accuracy was lower than experts’ and trainees’, but it outperformed all clinicians at delineating the placenta, IoU = 0.75 vs 0.69, 0.66, 0.59. The model was cross validated on 100 2nd trimester images from Barcelona, yielding an accuracy of 76%, IoU 0.68. In conclusion, we developed a model for automatic classification and segmentation of the placenta with consistent performance across different patient populations. It may be used for automated detection of placenta previa and enable future deep learning research in placental dysfunction.

Ensemble learning for fetal ultrasound and maternal–fetal data to predict mode of delivery after labor induction

“…The models’ parameters were learned using the training set; the prediction error for hyperparameter tuning and model selection was estimated using the validation set. The test set evaluated the generalization error for each of the final models, to avoid potential overfitting to the training set 17 , 20 .…”

“…CNNs are used in obstetric US for fetal weight estimation by measuring fetal biometry, identification of normal and abnormal anatomy, and detection and localization of structures and standard planes. This has already instituted clinical applications in fetal imaging, including echocardiography and neurosonography 15 , 20 . Recent literature findings by Kim et al’s 21 study group have developed a DL algorithm for automated measurement of BPD and HC, improving localization of fetal head shapes and caliper placement in later gestational ages.…”

Validity evidence supporting clinical skills assessment by artificial intelligence compared with trained clinician raters

“…This has implications for the interpretation of our study results as we may understate the potential value of AI‐based assessments. The robustness of the AIBA model may be improved by including more diverse training sets and by adjusting the model's hyperparameters to avoid overfitting to the training dataset 46–49 . However, the need for cross‐validation and very large datasets may ultimately hinder the accessibility and use of AI for assessment purposes, in particular, when compared with EBA that work after minimal rater instruction.…”

Evolving the Era of 5D Ultrasound? A Systematic Literature Review on the Applications for Artificial Intelligence Ultrasound Imaging in Obstetrics and Gynecology

“…Andreasen et al and Schilpzand et al presented an effective AI algorithm for placental localization, including heterogenous data through differences in sonographers' expertise [109], or using a previous established sweep protocol in low-resource settings [110]. It is known that early reduced placental volume is associated with small-for-gestational-age fetuses [111].…”

“…In summary, placental AI-based US diagnostic may propose a promising non-invasive, predictive tool to improve patient counseling and management to prevent adverse pregnancy outcomes. Reported limitations in applications arose from the difficulty of identifying the interface between the placenta and myometrium, especially in first trimester scans [113], and low accuracy rates in the assessment of posterior wall placentas [109]. Further research is necessary to identify the link between placental health and obstetric complications.…”