In this section
@article{Ma:2025:10.1016/j.media.2025.103902,
author = {Ma, X and Tao, Y and Zhang, Z and Zhang, Y and Wang, X and Zhang, S and Ji, Z and Zhang, Y and Chen, Q and Yang, G},
doi = {10.1016/j.media.2025.103902},
journal = {Med Image Anal},
title = {Test-time generative augmentation for medical image segmentation.},
url = {http://dx.doi.org/10.1016/j.media.2025.103902},
volume = {109},
year = {2025}
}
TY - JOUR
AB - Medical image segmentation is critical for clinical diagnosis, treatment planning, and monitoring, yet segmentation models often struggle with uncertainties stemming from occlusions, ambiguous boundaries, and variations in imaging devices. Traditional test-time augmentation (TTA) techniques typically rely on predefined geometric and photometric transformations, limiting their adaptability and effectiveness in complex medical scenarios. In this study, we introduced Test-Time Generative Augmentation (TTGA), a novel augmentation strategy specifically tailored for medical image segmentation at inference time. Different from conventional augmentation strategies that suffer from excessive randomness or limited flexibility, TTGA leverages a domain-fine-tuned generative model to produce contextually relevant and diverse augmentations tailored to the characteristics of each test image. Built upon diffusion model inversion, a masked null-text inversion method is proposed to enable region-specific augmentations during sampling. Furthermore, a dual denoising pathway is designed to balance precise identity preservation with controlled variability. We demonstrate the efficacy of our TTGA through extensive experiments across three distinct segmentation tasks spanning nine datasets. Our results consistently demonstrate that TTGA not only improves segmentation accuracy (with DSC gains ranging from 0.1 % to 2.3 % over the baseline) but also offers pixel-wise error estimation (with DSC gains ranging from 1.1 % to 29.0 % over the baseline). The source code and demonstration are available at: https://github.com/maxiao0234/TTGA.
AU - Ma,X
AU - Tao,Y
AU - Zhang,Z
AU - Zhang,Y
AU - Wang,X
AU - Zhang,S
AU - Ji,Z
AU - Zhang,Y
AU - Chen,Q
AU - Yang,G
DO - 10.1016/j.media.2025.103902
PY - 2025///
TI - Test-time generative augmentation for medical image segmentation.
T2 - Med Image Anal
UR - http://dx.doi.org/10.1016/j.media.2025.103902
UR - https://www.ncbi.nlm.nih.gov/pubmed/41389685
VL - 109
ER -
For enquiries about the MRI Physics Collective, please contact:
Mary Finnegan
Senior MR Physicist at the Imperial College Healthcare NHS Trust
Pete Lally
Assistant Professor in Magnetic Resonance (MR) Physics at Imperial College
Jan Sedlacik
MR Physicist at the Robert Steiner MR Unit, Hammersmith Hospital Campus