DiffSeg: A Segmentation Model for Skin Lesions Based on Diffusion Difference

TL;DR

DiffSeg is a novel diffusion-based segmentation model for skin lesions that leverages noise features, provides multiple outputs for uncertainty measurement, and refines results with DenseCRF, outperforming existing methods.

Contribution

The paper introduces DiffSeg, a diffusion difference-based segmentation model that captures uncertainty and improves accuracy through multi-output and DenseCRF integration.

Findings

Outperforms state-of-the-art U-Net methods on ISIC 2018 dataset

Provides uncertainty quantification via Generalized Energy Distance

Utilizes diffusion principles for noise feature extraction

Abstract

Weakly supervised medical image segmentation (MIS) using generative models is crucial for clinical diagnosis. However, the accuracy of the segmentation results is often limited by insufficient supervision and the complex nature of medical imaging. Existing models also only provide a single outcome, which does not allow for the measurement of uncertainty. In this paper, we introduce DiffSeg, a segmentation model for skin lesions based on diffusion difference which exploits diffusion model principles to ex-tract noise-based features from images with diverse semantic information. By discerning difference between these noise features, the model identifies diseased areas. Moreover, its multi-output capability mimics doctors' annotation behavior, facilitating the visualization of segmentation result consistency and ambiguity. Additionally, it quantifies output uncertainty using Generalized Energy Distance (GED), aiding interpretability and decision-making for physicians. Finally, the model integrates outputs through the Dense Conditional Random Field (DenseCRF) algorithm to refine the segmentation boundaries by considering inter-pixel correlations, which improves the accuracy and optimizes the segmentation results. We demonstrate the effectiveness of DiffSeg on the ISIC 2018 Challenge dataset, outperforming state-of-the-art U-Net-based methods.