Please use this identifier to cite or link to this item:
https://hdl.handle.net/2440/133384
Citations | ||
Scopus | Web of Science® | Altmetric |
---|---|---|
?
|
?
|
Type: | Conference paper |
Title: | Balanced-MixUp for highly imbalanced medical image classification |
Author: | Galdran, A. Carneiro, G. González Ballester, M.A. |
Citation: | Lecture Notes in Artificial Intelligence, 2021 / deBruijne, M., Cattin, P.C., Cotin, S., Padoy, N., Speidel, S., Zheng, Y., Essert, C. (ed./s), vol.12905, pp.323-333 |
Publisher: | Springer International Publishing |
Issue Date: | 2021 |
Series/Report no.: | Lecture Notes in Computer Science; 12905 |
ISBN: | 9783030872397 |
ISSN: | 0302-9743 1611-3349 |
Conference Name: | 24th International Conference on Medical Image Computing and Computer Assisted Intervention – MICCAI 2021 (27 Sep 2021 - 1 Oct 2021 : Strasbourg) |
Editor: | deBruijne, M. Cattin, P.C. Cotin, S. Padoy, N. Speidel, S. Zheng, Y. Essert, C. |
Statement of Responsibility: | Adrian Galdran, Gustavo Carneiro, Miguel A. González Ballester |
Abstract: | Highly imbalanced datasets are ubiquitous in medical image classification problems. In such problems, it is often the case that rare classes associated to less prevalent diseases are severely under-represented in labeled databases, typically resulting in poor performance of machine learning algorithms due to overfitting in the learning process. In this paper, we propose a novel mechanism for sampling training data based on the popular MixUp regularization technique, which we refer to as Balanced-MixUp. In short, Balanced-MixUp simultaneously performs regular (i.e., instance-based) and balanced (i.e., class-based) sampling of the training data. The resulting two sets of samples are then mixed-up to create a more balanced training distribution from which a neural network can effectively learn without incurring in heavily under-fitting the minority classes. We experiment with a highly imbalanced dataset of retinal images (55K samples, 5 classes) and a long-tail dataset of gastro-intestinal video frames (10K images, 23 classes), using two CNNs of varying representation capabilities. Experimental results demonstrate that applying Balanced-MixUp outperforms other conventional sampling schemes and loss functions specifically designed to deal with imbalanced data. |
Rights: | © Springer Nature Switzerland AG 2021. |
DOI: | 10.1007/978-3-030-87240-3_31 |
Grant ID: | http://purl.org/au-research/grants/arc/DP180103232 http://purl.org/au-research/grants/arc/FT190100525 |
Published version: | https://link.springer.com/chapter/10.1007%2F978-3-030-87240-3_31 |
Appears in Collections: | Computer Science publications |
Files in This Item:
There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.