DefCor-Net: Physics-Aware Ultrasound Deformation Correction

TL;DR

This paper introduces DefCor-Net, a deep neural network that uses real-time tissue property estimation and physics-based modeling to improve deformation correction in ultrasound images, enhancing anatomical accuracy for better diagnosis.

Contribution

The novel DefCor-Net combines physics-aware modeling with deep learning to perform pixel-wise deformation correction in ultrasound images, incorporating tissue stiffness estimation for improved accuracy.

Findings

Significantly improved deformation correction accuracy (Dice coefficient from 14.3 to 82.6).

Effective real-time tissue property estimation for anatomy-aware correction.

Validated on multiple volunteer datasets with consistent results.

Abstract

The recovery of morphologically accurate anatomical images from deformed ones is challenging in ultrasound (US) image acquisition, but crucial to accurate and consistent diagnosis, particularly in the emerging field of computer-assisted diagnosis. This article presents a novel anatomy-aware deformation correction approach based on a coarse-to-fine, multi-scale deep neural network (DefCor-Net). To achieve pixel-wise performance, DefCor-Net incorporates biomedical knowledge by estimating pixel-wise stiffness online using a U-shaped feature extractor. The deformation field is then computed using polynomial regression by integrating the measured force applied by the US probe. Based on real-time estimation of pixel-by-pixel tissue properties, the learning-based approach enables the potential for anatomy-aware deformation correction. To demonstrate the effectiveness of the proposed DefCor-Net, images recorded at multiple locations on forearms and upper arms of six volunteers are used to train and validate DefCor-Net. The results demonstrate that DefCor-Net can significantly improve the accuracy of deformation correction to recover the original geometry (Dice Coefficient: from $14.3\pm20.9$ to $82.6\pm12.1$ when the force is $6N$).