A semi-supervised domain adaptive medical image segmentation method based on dual-level multi-scale alignment

Abstract In the actual image segmentation tasks in the medical field, the phenomenon of limited labeled data accompanied by domain shifts often occurs and such domain shifts may exist in homologous or even heterologous data. In the study, a novel method was proposed to deal with this challenging phe...

Full description

Saved in:
Bibliographic Details
Main Authors: Hualing Li, Yaodan Wang, Yan Qiang
Format: Article
Language:English
Published: Nature Portfolio 2025-03-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-025-93824-6
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850023390877843456
author Hualing Li
Yaodan Wang
Yan Qiang
author_facet Hualing Li
Yaodan Wang
Yan Qiang
author_sort Hualing Li
collection DOAJ
description Abstract In the actual image segmentation tasks in the medical field, the phenomenon of limited labeled data accompanied by domain shifts often occurs and such domain shifts may exist in homologous or even heterologous data. In the study, a novel method was proposed to deal with this challenging phenomenon. Firstly, a model was trained with labeled data in source and target domains so as to adapt to unlabeled data. Then, the alignment at two main levels was realized. At the style level, based on multi-scale stylistic features, the alignment of unlabeled target images was maximized and unlabeled target image features were enhanced. At the inter-domain level, the similarity of the category centroids between target domain data and mixed image data was also maximized. Additionally, a fused supervised loss and alignment loss computation method was proposed. In validation experiments, two cross-domain medical image datasets were constructed: homologous and heterologous datasets. Experimental results showed that the proposed method had the more advantageous comprehensive performance than common semi-supervised and domain adaptation methods.
format Article
id doaj-art-556e81be6f0840cc8df8cead014c7361
institution DOAJ
issn 2045-2322
language English
publishDate 2025-03-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-556e81be6f0840cc8df8cead014c73612025-08-20T03:01:23ZengNature PortfolioScientific Reports2045-23222025-03-0115111510.1038/s41598-025-93824-6A semi-supervised domain adaptive medical image segmentation method based on dual-level multi-scale alignmentHualing Li0Yaodan Wang1Yan Qiang2School of Software, North University of ChinaSchool of Software, North University of ChinaSchool of Software, North University of ChinaAbstract In the actual image segmentation tasks in the medical field, the phenomenon of limited labeled data accompanied by domain shifts often occurs and such domain shifts may exist in homologous or even heterologous data. In the study, a novel method was proposed to deal with this challenging phenomenon. Firstly, a model was trained with labeled data in source and target domains so as to adapt to unlabeled data. Then, the alignment at two main levels was realized. At the style level, based on multi-scale stylistic features, the alignment of unlabeled target images was maximized and unlabeled target image features were enhanced. At the inter-domain level, the similarity of the category centroids between target domain data and mixed image data was also maximized. Additionally, a fused supervised loss and alignment loss computation method was proposed. In validation experiments, two cross-domain medical image datasets were constructed: homologous and heterologous datasets. Experimental results showed that the proposed method had the more advantageous comprehensive performance than common semi-supervised and domain adaptation methods.https://doi.org/10.1038/s41598-025-93824-6Medical image segmentationSemi-supervisedDomain adaptiveMulti-scale alignment
spellingShingle Hualing Li
Yaodan Wang
Yan Qiang
A semi-supervised domain adaptive medical image segmentation method based on dual-level multi-scale alignment
Scientific Reports
Medical image segmentation
Semi-supervised
Domain adaptive
Multi-scale alignment
title A semi-supervised domain adaptive medical image segmentation method based on dual-level multi-scale alignment
title_full A semi-supervised domain adaptive medical image segmentation method based on dual-level multi-scale alignment
title_fullStr A semi-supervised domain adaptive medical image segmentation method based on dual-level multi-scale alignment
title_full_unstemmed A semi-supervised domain adaptive medical image segmentation method based on dual-level multi-scale alignment
title_short A semi-supervised domain adaptive medical image segmentation method based on dual-level multi-scale alignment
title_sort semi supervised domain adaptive medical image segmentation method based on dual level multi scale alignment
topic Medical image segmentation
Semi-supervised
Domain adaptive
Multi-scale alignment
url https://doi.org/10.1038/s41598-025-93824-6
work_keys_str_mv AT hualingli asemisuperviseddomainadaptivemedicalimagesegmentationmethodbasedonduallevelmultiscalealignment
AT yaodanwang asemisuperviseddomainadaptivemedicalimagesegmentationmethodbasedonduallevelmultiscalealignment
AT yanqiang asemisuperviseddomainadaptivemedicalimagesegmentationmethodbasedonduallevelmultiscalealignment
AT hualingli semisuperviseddomainadaptivemedicalimagesegmentationmethodbasedonduallevelmultiscalealignment
AT yaodanwang semisuperviseddomainadaptivemedicalimagesegmentationmethodbasedonduallevelmultiscalealignment
AT yanqiang semisuperviseddomainadaptivemedicalimagesegmentationmethodbasedonduallevelmultiscalealignment