Semi-Supervised Segmentation via Embedding Matching

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Xie, Weiyi, Willems, Nathalie, Lessmann, Nikolas, Gibbons, Tom, De Massari, Daniele
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916313606127616
author Xie, Weiyi
Willems, Nathalie
Lessmann, Nikolas
Gibbons, Tom
De Massari, Daniele
author_facet Xie, Weiyi
Willems, Nathalie
Lessmann, Nikolas
Gibbons, Tom
De Massari, Daniele
contents Deep convolutional neural networks are widely used in medical image segmentation but require many labeled images for training. Annotating three-dimensional medical images is a time-consuming and costly process. To overcome this limitation, we propose a novel semi-supervised segmentation method that leverages mostly unlabeled images and a small set of labeled images in training. Our approach involves assessing prediction uncertainty to identify reliable predictions on unlabeled voxels from the teacher model. These voxels serve as pseudo-labels for training the student model. In voxels where the teacher model produces unreliable predictions, pseudo-labeling is carried out based on voxel-wise embedding correspondence using reference voxels from labeled images. We applied this method to automate hip bone segmentation in CT images, achieving notable results with just 4 CT scans. The proposed approach yielded a Hausdorff distance with 95th percentile (HD95) of 3.30 and IoU of 0.929, surpassing existing methods achieving HD95 (4.07) and IoU (0.927) at their best.
format Preprint
id arxiv_https___arxiv_org_abs_2407_04638
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Semi-Supervised Segmentation via Embedding Matching
Xie, Weiyi
Willems, Nathalie
Lessmann, Nikolas
Gibbons, Tom
De Massari, Daniele
Computer Vision and Pattern Recognition
I.5.4; I.4.6; I.2.10
Deep convolutional neural networks are widely used in medical image segmentation but require many labeled images for training. Annotating three-dimensional medical images is a time-consuming and costly process. To overcome this limitation, we propose a novel semi-supervised segmentation method that leverages mostly unlabeled images and a small set of labeled images in training. Our approach involves assessing prediction uncertainty to identify reliable predictions on unlabeled voxels from the teacher model. These voxels serve as pseudo-labels for training the student model. In voxels where the teacher model produces unreliable predictions, pseudo-labeling is carried out based on voxel-wise embedding correspondence using reference voxels from labeled images. We applied this method to automate hip bone segmentation in CT images, achieving notable results with just 4 CT scans. The proposed approach yielded a Hausdorff distance with 95th percentile (HD95) of 3.30 and IoU of 0.929, surpassing existing methods achieving HD95 (4.07) and IoU (0.927) at their best.
title Semi-Supervised Segmentation via Embedding Matching
topic Computer Vision and Pattern Recognition
I.5.4; I.4.6; I.2.10
url https://arxiv.org/abs/2407.04638