Deep Learning Analysis of Prenatal Ultrasound for Identification of Ventriculomegaly

2025-11-11

AI Summary

This study addresses the early automated detection of fetal ventriculomegaly in prenatal ultrasound. We propose a fine-tuning approach based on the self-supervised pretrained ultrasound foundation model USF-MAE—its first application to binary classification of fetal ventricular dilation. The model employs a Masked Autoencoder architecture built upon Vision Transformer (ViT), pretrained on large-scale unlabeled ultrasound data, and integrates Eigen-CAM for clinically interpretable visualization. Evaluated via 5-fold cross-validation and an independent test set, it achieves an F1 score of 91.76% (CV) and 91.78% (test), with 97.24% accuracy—significantly outperforming baselines including VGG-19, ResNet-50, and ViT-B/16. This work demonstrates the effective transferability of USF-MAE to fetal neurosonographic diagnosis and establishes a high-accuracy, interpretable AI tool for early risk assessment of chromosomal abnormalities and genetic syndromes in prenatal care.

Technology Category

• Computer Vision: Other Foundations of Computer Vision
• Machine Learning: Unsupervised & Self-Supervised Learning
• Natural Language Processing: Safety and Robustness

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

The proposed study aimed to develop a deep learning model capable of detecting ventriculomegaly on prenatal ultrasound images. Ventriculomegaly is a prenatal condition characterized by dilated cerebral ventricles of the fetal brain and is important to diagnose early, as it can be associated with an increased risk for fetal aneuploidies and/or underlying genetic syndromes. An Ultrasound Self-Supervised Foundation Model with Masked Autoencoding (USF-MAE), recently developed by our group, was fine-tuned for a binary classification task to distinguish fetal brain ultrasound images as either normal or showing ventriculomegaly. The USF-MAE incorporates a Vision Transformer encoder pretrained on more than 370,000 ultrasound images from the OpenUS-46 corpus. For this study, the pretrained encoder was adapted and fine-tuned on a curated dataset of fetal brain ultrasound images to optimize its performance for ventriculomegaly detection. Model evaluation was conducted using 5-fold cross-validation and an independent test cohort, and performance was quantified using accuracy, precision, recall, specificity, F1-score, and area under the receiver operating characteristic curve (AUC). The proposed USF-MAE model reached an F1-score of 91.76% on the 5-fold cross-validation and 91.78% on the independent test set, with much higher scores than those obtained by the baseline models by 19.37% and 16.15% compared to VGG-19, 2.31% and 2.56% compared to ResNet-50, and 5.03% and 11.93% compared to ViT-B/16, respectively. The model also showed a high mean test precision of 94.47% and an accuracy of 97.24%. The Eigen-CAM (Eigen Class Activation Map) heatmaps showed that the model was focusing on the ventricle area for the diagnosis of ventriculomegaly, which has explainability and clinical plausibility.

Problem

• Developing deep learning model to detect ventriculomegaly in prenatal ultrasound images
• Fine-tuning self-supervised foundation model for fetal brain abnormality classification
• Automating diagnosis of dilated cerebral ventricles using vision transformer architecture

Innovation

• Fine-tuned self-supervised model for ventriculomegaly classification
• Used Vision Transformer pretrained on large ultrasound dataset
• Achieved high accuracy with explainable ventricle-focused heatmaps