Metrics to Quantify Global Consistency in Synthetic Medical Images

TL;DR

This paper introduces two novel metrics to quantify the global consistency of synthetic medical images, ensuring they are biologically plausible and coherent, which is crucial for medical applications.

Contribution

The work presents explicit and implicit image attribute prediction methods to measure global consistency, filling a gap left by existing image quality metrics.

Findings

Explicit attribute prediction effectively distinguishes consistent from inconsistent images.

Implicit feature similarity is useful when labeled data is scarce.

Metrics outperform traditional measures like FID in assessing biological plausibility.

Abstract

Image synthesis is increasingly being adopted in medical image processing, for example for data augmentation or inter-modality image translation. In these critical applications, the generated images must fulfill a high standard of biological correctness. A particular requirement for these images is global consistency, i.e an image being overall coherent and structured so that all parts of the image fit together in a realistic and meaningful way. Yet, established image quality metrics do not explicitly quantify this property of synthetic images. In this work, we introduce two metrics that can measure the global consistency of synthetic images on a per-image basis. To measure the global consistency, we presume that a realistic image exhibits consistent properties, e.g., a person's body fat in a whole-body MRI, throughout the depicted object or scene. Hence, we quantify global consistency by predicting and comparing explicit attributes of images on patches using supervised trained neural networks. Next, we adapt this strategy to an unlabeled setting by measuring the similarity of implicit image features predicted by a self-supervised trained network. Our results demonstrate that predicting explicit attributes of synthetic images on patches can distinguish globally consistent from inconsistent images. Implicit representations of images are less sensitive to assess global consistency but are still serviceable when labeled data is unavailable. Compared to established metrics, such as the FID, our method can explicitly measure global consistency on a per-image basis, enabling a dedicated analysis of the biological plausibility of single synthetic images.