DiffSegLung: Diffusion Radiomic Distillation for Unsupervised Lung Pathology Segmentation

TL;DR

DiffSegLung introduces an unsupervised framework leveraging diffusion radiomic distillation and physics-based descriptors to improve lung pathology segmentation in CT scans without annotations.

Contribution

It proposes a novel unsupervised method combining radiomic descriptors and diffusion models for lung pathology segmentation.

Findings

Outperforms unsupervised baselines across four pathology classes.

Enhances segmentation accuracy and boundary refinement.

Improves generation fidelity over prior diffusion models.

Abstract

Unsupervised segmentation of pulmonary pathologies in CT remains an open challenge due to the absence of annotated multi pathology cohorts and the failure of existing diffusion-based methods to exploit the quantitative Hounsfield Unit (HU) signal that physically distinguishes tissue classes. To address this, we propose DiffSegLung,a framework that introduces Diffusion Radiomic Distillation, in which handcrafted radiomic descriptors serve as a physics grounded teacher to shape the bottleneck of a 3D diffusion U-Net via a contrastive objective, transferring pathology discriminative structure into the learned representation without any annotations. At inference, the teacher is discarded and multitimestep bottleneck features are clustered by a Gaussian Mixture Model with HU-guided label assignment, followed by Sobel Diffusion Fusion for boundary refinement. Evaluated on 190 expert annotated axial slices drawn from four heterogeneous CT cohorts, Diff-SegLung improves segmentation across all four pathology classes over unsupervised baselines and improves generation fidelity over prior CT diffusion models.