From Variability To Accuracy: Conditional Bernoulli Diffusion Models with Consensus-Driven Correction for Thin Structure Segmentation

TL;DR

This paper introduces a novel consensus-driven correction framework using conditional Bernoulli diffusion models to improve the accuracy and continuity of orbital bone segmentation in facial CT images, especially in ambiguous thin structures.

Contribution

The proposed method leverages multiple diffusion model outputs and a consensus correction to enhance segmentation accuracy in challenging regions, automating manual correction processes.

Findings

Outperforms existing segmentation methods in ambiguous regions

Significantly improves recall for thin structures

Automates correction process for surgical planning

Abstract

Accurate segmentation of orbital bones in facial computed tomography (CT) images is essential for the creation of customized implants for reconstruction of defected orbital bones, particularly challenging due to the ambiguous boundaries and thin structures such as the orbital medial wall and orbital floor. In these ambiguous regions, existing segmentation approaches often output disconnected or under-segmented results. We propose a novel framework that corrects segmentation results by leveraging consensus from multiple diffusion model outputs. Our approach employs a conditional Bernoulli diffusion model trained on diverse annotation patterns per image to generate multiple plausible segmentations, followed by a consensus-driven correction that incorporates position proximity, consensus level, and gradient direction similarity to correct challenging regions. Experimental results demonstrate that our method outperforms existing methods, significantly improving recall in ambiguous regions while preserving the continuity of thin structures. Furthermore, our method automates the manual process of segmentation result correction and can be applied to image-guided surgical planning and surgery.