Diffusion as Sound Propagation: Physics-inspired Model for Ultrasound Image Generation

Marina Domínguez,Yordanka Velikova,Nassir Navab,Mohammad Farid Azampour

2024-07-08

Abstract:Deep learning (DL) methods typically require large datasets to effectively learn data distributions. However, in the medical field, data is often limited in quantity, and acquiring labeled data can be costly. To mitigate this data scarcity, data augmentation techniques are commonly employed. Among these techniques, generative models play a pivotal role in expanding datasets. However, when it comes to ultrasound (US) imaging, the authenticity of generated data often diminishes due to the oversight of ultrasound physics. We propose a novel approach to improve the quality of generated US images by introducing a physics-based diffusion model that is specifically designed for this image modality. The proposed model incorporates an US-specific scheduler scheme that mimics the natural behavior of sound wave propagation in ultrasound imaging. Our analysis demonstrates how the proposed method aids in modeling the attenuation dynamics in US imaging. We present both qualitative and quantitative results based on standard generative model metrics, showing that our proposed method results in overall more plausible images. Our code is available at <a class="link-external link-https" href="https://github.com/marinadominguez/diffusion-for-us-images" rel="external noopener nofollow">this https URL</a>

Image and Video Processing,Computer Vision and Pattern Recognition

What problem does this paper attempt to address?

The paper attempts to address the issue of poor image quality in ultrasound image generation due to existing generative models neglecting the physical characteristics of ultrasound. Specifically, the paper points out that in the medical field, the amount of annotated data is limited and costly to obtain, so data augmentation techniques (especially generative models) are widely used to expand datasets. However, for ultrasound imaging, the authenticity of generated images is often reduced due to a lack of consideration of ultrasound physical properties.

To improve the quality of generated ultrasound images, the paper proposes a physics-based diffusion model specifically designed for the ultrasound image modality. This model introduces an ultrasound-specific scheduling scheme that simulates the natural behavior of sound wave propagation in ultrasound imaging, with particular attention to the attenuation dynamics of sound waves in tissues. In this way, the model can generate more realistic ultrasound images.

The main contributions of the paper include:
1. Proposing a new noise scheduler (B-Maps) that simulates the attenuation of sound waves as they return to the ultrasound receiver.
2. Considering depth-dependent changes in ultrasound resolution and giving more emphasis to areas near the probe (where image details and internal structures are clearer and more reliable).
3. Validating the effectiveness of the proposed method through qualitative and quantitative experiments, with results showing that the generated images perform better on multiple standard generative model metrics.

In summary, the paper aims to improve the realism of generated ultrasound images by introducing a physics-inspired diffusion model, thereby alleviating the problem of data scarcity and enhancing the quality of medical image generation.