CSEval: A Framework for Evaluating Clinical Semantics in Text-to-Image Generation

TL;DR

CSEval is a novel framework that uses language models to evaluate the clinical semantic accuracy of images generated from medical text prompts, addressing a gap in existing evaluation metrics.

Contribution

The paper introduces CSEval, a new framework that assesses clinical semantic alignment in medical text-to-image generation using language models, improving evaluation accuracy.

Findings

CSEval detects semantic inconsistencies missed by traditional metrics.

CSEval correlates well with expert clinical judgment.

Provides a scalable, clinically meaningful evaluation method.

Abstract

Text-to-image generation has been increasingly applied in medical domains for various purposes such as data augmentation and education. Evaluating the quality and clinical reliability of these generated images is essential. However, existing methods mainly assess image realism or diversity, while failing to capture whether the generated images reflect the intended clinical semantics, such as anatomical location and pathology. In this study, we propose the Clinical Semantics Evaluator (CSEval), a framework that leverages language models to assess clinical semantic alignment between the generated images and their conditioning prompts. Our experiments show that CSEval identifies semantic inconsistencies overlooked by other metrics and correlates with expert judgment. CSEval provides a scalable and clinically meaningful complement to existing evaluation methods, supporting the safe adoption of generative models in healthcare.