Projects per year
Abstract
Automated pathology segmentation remains a valuable diagnostic tool in clinical practice. However, collecting training data is challenging. Semi-supervised approaches by combining labelled and unlabelled data can offer a solution to data scarcity. An approach to semisupervised learning relies on reconstruction objectives (as self-supervision objectives) that learns in a joint fashion suitable representations for the task. Here, we propose Anatomy-Pathology Disentanglement Network (APD-Net), a pathology segmentation model that attempts to learn jointly for the first time: disentanglement of anatomy, modality, and pathology. The model is trained in a semi-supervised fashion with new reconstruction losses directly aiming to improve pathology segmentation with limited annotations. In addition, a joint optimization strategy is proposed to fully take advantage of the available annotations. We evaluate our methods with two private cardiac infarction segmentation datasets with LGE-MRI scans. APD-Net can perform pathology segmentation with few annotations, maintain performance with different amounts of supervision, and outperform related deep learning methods.
Original language | English |
---|---|
Title of host publication | DART: MICCAI Workshop on Domain Adaptation and Representation Transfer |
Subtitle of host publication | Domain Adaptation and Representation Transfer, and Distributed and Collaborative Learning |
Publisher | Springer |
Pages | 62-72 |
ISBN (Electronic) | 978-3-030-60548-3 |
ISBN (Print) | 978-3-030-60547-6 |
DOIs | |
Publication status | E-pub ahead of print - 26 Sept 2020 |
Event | MICCAI 2020 : DART Workshop - Duration: 8 Sept 2020 → 8 Sept 2020 |
Publication series
Name | Lecture Notes in Computer Science |
---|---|
Publisher | Springer Link |
Volume | 12444 |
ISSN (Print) | 0302-9743 |
ISSN (Electronic) | 1611-3349 |
Conference
Conference | MICCAI 2020 |
---|---|
Period | 8/09/20 → 8/09/20 |
Keywords / Materials (for Non-textual outputs)
- pathology segmentation
- disentangled representations
- semi- supervised learning
Fingerprint
Dive into the research topics of 'Semi-supervised Pathology Segmentation with Disentangled Representations'. Together they form a unique fingerprint.Projects
- 1 Active
-
Canon Medical / RAEng Senior Research Fellow in Healthcare AI
Tsaftaris, S. (Principal Investigator)
Canon Medical Research Europe Limited
31/03/19 → 30/06/26
Project: Research