DCD: A Semantic Segmentation Model for Fetal Ultrasound Four-Chamber View

TL;DR

This paper introduces DCD, a deep learning model that improves fetal heart structure segmentation in ultrasound images by using advanced modules for multi-scale feature extraction and attention, aiding early diagnosis of congenital heart disease.

Contribution

The novel DCD model combines Dense ASPP and CBAM modules to enhance segmentation accuracy in fetal ultrasound images, addressing challenges like noise and anatomical variability.

Findings

DCD outperforms existing models in segmentation accuracy.

The model demonstrates robustness across different gestational stages.

Enhanced feature extraction improves clinical assessment.

Abstract

Accurate segmentation of anatomical structures in the apical four-chamber (A4C) view of fetal echocardiography is essential for early diagnosis and prenatal evaluation of congenital heart disease (CHD). However, precise segmentation remains challenging due to ultrasound artifacts, speckle noise, anatomical variability, and boundary ambiguity across different gestational stages. To reduce the workload of sonographers and enhance segmentation accuracy, we propose DCD, an advanced deep learning-based model for automatic segmentation of key anatomical structures in the fetal A4C view. Our model incorporates a Dense Atrous Spatial Pyramid Pooling (Dense ASPP) module, enabling superior multi-scale feature extraction, and a Convolutional Block Attention Module (CBAM) to enhance adaptive feature representation. By effectively capturing both local and global contextual information, DCD achieves precise and robust segmentation, contributing to improved prenatal cardiac assessment.