XAI-CLIP: ROI-Guided Perturbation Framework for Explainable Medical Image Segmentation in Multimodal Vision-Language Models

TL;DR

XAI-CLIP introduces an ROI-guided perturbation framework that leverages multimodal vision-language models to produce clearer, more accurate, and computationally efficient explanations for medical image segmentation.

Contribution

It presents a novel ROI-guided perturbation method utilizing vision-language embeddings to improve interpretability and reduce computation in medical image segmentation explanations.

Findings

Achieves up to 60% reduction in runtime.

Improves dice score by 44.6%.

Increases IoU by 96.7% for explanations.

Abstract

Medical image segmentation is a critical component of clinical workflows, enabling accurate diagnosis, treatment planning, and disease monitoring. However, despite the superior performance of transformer-based models over convolutional architectures, their limited interpretability remains a major obstacle to clinical trust and deployment. Existing explainable artificial intelligence (XAI) techniques, including gradient-based saliency methods and perturbation-based approaches, are often computationally expensive, require numerous forward passes, and frequently produce noisy or anatomically irrelevant explanations. To address these limitations, we propose XAI-CLIP, an ROI-guided perturbation framework that leverages multimodal vision-language model embeddings to localize clinically meaningful anatomical regions and guide the explanation process. By integrating language-informed region localization with medical image segmentation and applying targeted, region-aware perturbations, the proposed method generates clearer, boundary-aware saliency maps while substantially reducing computational overhead. Experiments conducted on the FLARE22 and CHAOS datasets demonstrate that XAI-CLIP achieves up to a 60\% reduction in runtime, a 44.6\% improvement in dice score, and a 96.7\% increase in Intersection-over-Union for occlusion-based explanations compared to conventional perturbation methods. Qualitative results further confirm cleaner and more anatomically consistent attribution maps with fewer artifacts, highlighting that the incorporation of multimodal vision-language representations into perturbation-based XAI frameworks significantly enhances both interpretability and efficiency, thereby enabling transparent and clinically deployable medical image segmentation systems.