Deep learning from small labeled datasets applied to medical image analysis Book Section
| Authors: | Veeraraghavan, H.; Jiang, J. |
| Editors: | El-Baz, A. S.; Suri, J. S. |
| Article/Chapter Title: | Deep learning from small labeled datasets applied to medical image analysis |
| Abstract: | Deep learning has been a hot topic and applied in various applications. However, a crucial requirement of deep learning in medical image applications is the ability to produce generalizable learning from small and often heterogeneous image sets. In this chapter, we will present some deep learning approaches for learning from small datasets. These approaches use the following idea: leverage information from a different related modality for learning. While approaches exist even for learning generalizable models from the same modality, we particularly focus on the problem where learning is done by using a different imaging modality such as computed tomography (CT) with MRI, T1-weighted MRI with FLAIR MRI, CT with positron emission tomography, etc. Learning using from different modalities is called cross-modality learning. © 2021 Elsevier Inc. All rights reserved. |
| Keywords: | mri; computed tomography; deep learning; cross-modality learning; adversarial learning techniques |
| Book Title: | State of the Art in Neural Networks and Their Applications: Volume 1 |
| ISBN: | 978-0-12-819740-0 |
| Publisher: | Academic Press |
| Publication Place: | London, United Kingdom |
| Date Published: | 2021-01-01 |
| Start Page: | 279 |
| End Page: | 291 |
| Language: | English |
| DOI: | 10.1016/b978-0-12-819740-0.00014-0 |
| PROVIDER: | scopus |
| DOI/URL: | |
| Notes: | Book Chapter -- Export Date: 1 April 2022 -- Source: Scopus; Chapter 14 of book. |
Altmetric
Citation Impact
BMJ Impact Analytics
MSK Authors
-
159Veeraraghavan -
78Jiang
Related MSK Work