Assessing the capacity of a denoising diffusion probabilistic model to reproduce spatial context

TL;DR

This study systematically evaluates how well denoising diffusion probabilistic models (DDPMs) can learn and reproduce spatial context in medical imaging, revealing their potential advantages over GANs for data augmentation.

Contribution

First comprehensive assessment of DDPMs' ability to learn spatial context in medical images using stochastic context models and quantitative analysis.

Findings

DDPMs can generate contextually accurate images interpolated between training samples.

DDPMs outperform GANs in reproducing spatial context relevant to medical imaging.

Results suggest DDPMs are promising for data augmentation in medical imaging applications.

Abstract

Diffusion models have emerged as a popular family of deep generative models (DGMs). In the literature, it has been claimed that one class of diffusion models -- denoising diffusion probabilistic models (DDPMs) -- demonstrate superior image synthesis performance as compared to generative adversarial networks (GANs). To date, these claims have been evaluated using either ensemble-based methods designed for natural images, or conventional measures of image quality such as structural similarity. However, there remains an important need to understand the extent to which DDPMs can reliably learn medical imaging domain-relevant information, which is referred to as `spatial context' in this work. To address this, a systematic assessment of the ability of DDPMs to learn spatial context relevant to medical imaging applications is reported for the first time. A key aspect of the studies is the use of stochastic context models (SCMs) to produce training data. In this way, the ability of the DDPMs to reliably reproduce spatial context can be quantitatively assessed by use of post-hoc image analyses. Error-rates in DDPM-generated ensembles are reported, and compared to those corresponding to a modern GAN. The studies reveal new and important insights regarding the capacity of DDPMs to learn spatial context. Notably, the results demonstrate that DDPMs hold significant capacity for generating contextually correct images that are `interpolated' between training samples, which may benefit data-augmentation tasks in ways that GANs cannot.