Deep Image Translation for Enhancing Simulated Ultrasound Images

TL;DR

This paper presents a deep learning method that enhances the quality of simulated ultrasound images in real-time, combining adversarial training, anatomical segmentation, and acoustic information to improve realism without increasing computational load.

Contribution

It introduces a novel image translation framework that improves ultrasound simulation quality by incorporating anatomical and acoustic data, maintaining constant computation time.

Findings

7.2% improvement in Fréchet Inception Distance

8.9% improvement in patch-based Kullback-Leibler divergence

Enhanced ultrasound image realism with real-time performance

Abstract

Ultrasound simulation based on ray tracing enables the synthesis of highly realistic images. It can provide an interactive environment for training sonographers as an educational tool. However, due to high computational demand, there is a trade-off between image quality and interactivity, potentially leading to sub-optimal results at interactive rates. In this work we introduce a deep learning approach based on adversarial training that mitigates this trade-off by improving the quality of simulated images with constant computation time. An image-to-image translation framework is utilized to translate low quality images into high quality versions. To incorporate anatomical information potentially lost in low quality images, we additionally provide segmentation maps to image translation. Furthermore, we propose to leverage information from acoustic attenuation maps to better preserve acoustic shadows and directional artifacts, an invaluable feature for ultrasound image interpretation. The proposed method yields an improvement of 7.2% in Fr\'{e}chet Inception Distance and 8.9% in patch-based Kullback-Leibler divergence.