Abstract
Unsupervised domain adaptation (UDA) has become increasingly popular in imaging-based diagnosis due to the challenge of labeling a large number of datasets in target domains. Without labeled data, well-trained deep learning models in a source domain may not perform well when applied to a target domain. UDA allows for the use of large-scale datasets from various domains for model deployment, but it can face difficulties in performing adaptive feature extraction when dealing with unlabeled data in an unseen target domain. To address this, we propose an advanced test-time fine-tuning UDA framework designed to better utilize the latent features of datasets in the unseen target domain by fine-tuning the model itself during diagnosis. Our proposed framework is based on an auto-encoder-based network architecture that fine-tunes the model itself. This allows our framework to learn knowledge specific to the unseen target domain during the fine-tuning phase. In order to further optimize our framework for the unseen target domain, we introduce a re-initialization module that injects randomness into network parameters. This helps the framework to converge to a local minimum that is better-suited for the target domain, allowing for improved performance in domain adaptation tasks. To evaluate our framework, we carried out experiments on UDA segmentation tasks using breast cancer datasets acquired from multiple domains. Our experimental results demonstrated that our framework achieved state-of-the-art performance, outperforming other competing UDA models, in segmenting breast cancer on ultrasound images from an unseen domain, which supports its clinical potential for improving breast cancer diagnosis.
Original language | English |
---|---|
Title of host publication | Medical Image Computing and Computer Assisted Intervention – MICCAI 2023 - 26th International Conference, Proceedings |
Editors | Hayit Greenspan, Hayit Greenspan, Anant Madabhushi, Parvin Mousavi, Septimiu Salcudean, James Duncan, Tanveer Syeda-Mahmood, Russell Taylor |
Publisher | Springer Science and Business Media Deutschland GmbH |
Pages | 539-550 |
Number of pages | 12 |
ISBN (Print) | 9783031439063 |
DOIs | |
State | Published - 2023 |
Event | 26th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2023 - Vancouver, Canada Duration: 8 Oct 2023 → 12 Oct 2023 |
Publication series
Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
---|---|
Volume | 14220 LNCS |
ISSN (Print) | 0302-9743 |
ISSN (Electronic) | 1611-3349 |
Conference
Conference | 26th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2023 |
---|---|
Country/Territory | Canada |
City | Vancouver |
Period | 8/10/23 → 12/10/23 |
Bibliographical note
Publisher Copyright:© 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.
Keywords
- Breast Cancer
- Segmentation
- Test-Time Tuning
- Ultrasound Imaging
- Unsupervised Domain Adaptation