Weakly-Supervised Lung Nodule Segmentation via Training-Free Guidance of 3D Rectified Flow

TL;DR

This paper introduces a training-free, weakly-supervised lung nodule segmentation method using pretrained 3D rectified flow models, requiring only minimal fine-tuning and achieving improved results on LUNA16.

Contribution

It presents a novel approach combining pretrained generative models with image-level labels for accurate lung nodule segmentation without retraining the generative model.

Findings

Improved segmentation quality over baseline methods.

Consistent detection of lung nodules of various sizes and shapes.

Effective use of training-free guidance with pretrained models.

Abstract

Dense annotations, such as segmentation masks, are expensive and time-consuming to obtain, especially for 3D medical images where expert voxel-wise labeling is required. Weakly supervised approaches aim to address this limitation, but often rely on attribution-based methods that struggle to accurately capture small structures such as lung nodules. In this paper, we propose a weakly-supervised segmentation method for lung nodules by combining pretrained state-of-the-art rectified flow and predictor models in a plug-and-play manner. Our approach uses training-free guidance of a 3D rectified flow model, requiring only fine-tuning of the predictor using image-level labels and no retraining of the generative model. The proposed method produces improved-quality segmentations for two separate predictors, consistently detecting lung nodules of varying size and shapes. Experiments on LUNA16 demonstrate improvements over baseline methods, highlighting the potential of generative foundation models as tools for weakly supervised 3D medical image segmentation.