Weakly Supervised Localisation for Fetal Ultrasound Images

TL;DR

This study develops a weakly supervised CNN approach with soft proposal layers to detect and localize fetal anatomical regions in 2D ultrasound images using only image-level labels, achieving high accuracy.

Contribution

It introduces a novel weakly supervised method combining CNNs and soft proposal layers for fetal ultrasound localization without segmentation labels.

Findings

Achieved 90% average accuracy in detecting fetal regions

Proposal maps effectively highlight relevant anatomical structures

Method enables future fetal pose and organ segmentation tasks

Abstract

This paper addresses the task of detecting and localising fetal anatomical regions in 2D ultrasound images, where only image-level labels are present at training, i.e. without any localisation or segmentation information. We examine the use of convolutional neural network architectures coupled with soft proposal layers. The resulting network simultaneously performs anatomical region detection (classification) and localisation tasks. We generate a proposal map describing the attention of the network for a particular class. The network is trained on 85,500 2D fetal Ultrasound images and their associated labels. Labels correspond to six anatomical regions: head, spine, thorax, abdomen, limbs, and placenta. Detection achieves an average accuracy of 90\% on individual regions, and show that the proposal maps correlate well with relevant anatomical structures. This work presents itself as a powerful and essential step towards subsequent tasks such as fetal position and pose estimation, organ-specific segmentation, or image-guided navigation. Code and additional material is available at https://ntoussaint.github.io/fetalnav