The effect of the loss on generalization: empirical study on synthetic
lung nodule data
lung nodule data
File(s)
Accepted version
Author(s)
Baltatzis, Vasileios
Folgoc, Loic Le
Ellis, Sam
Manzanera, Octavio E Martinez
Bintsi, Kyriaki-Margarita
Type
Conference Paper
Abstract
Convolutional Neural Networks (CNNs) are widely used for image classification
in a variety of fields, including medical imaging. While most studies deploy
cross-entropy as the loss function in such tasks, a growing number of
approaches have turned to a family of contrastive learning-based losses. Even
though performance metrics such as accuracy, sensitivity and specificity are
regularly used for the evaluation of CNN classifiers, the features that these
classifiers actually learn are rarely identified and their effect on the
classification performance on out-of-distribution test samples is
insufficiently explored. In this paper, motivated by the real-world task of
lung nodule classification, we investigate the features that a CNN learns when
trained and tested on different distributions of a synthetic dataset with
controlled modes of variation. We show that different loss functions lead to
different features being learned and consequently affect the generalization
ability of the classifier on unseen data. This study provides some important
insights into the design of deep learning solutions for medical imaging tasks.
in a variety of fields, including medical imaging. While most studies deploy
cross-entropy as the loss function in such tasks, a growing number of
approaches have turned to a family of contrastive learning-based losses. Even
though performance metrics such as accuracy, sensitivity and specificity are
regularly used for the evaluation of CNN classifiers, the features that these
classifiers actually learn are rarely identified and their effect on the
classification performance on out-of-distribution test samples is
insufficiently explored. In this paper, motivated by the real-world task of
lung nodule classification, we investigate the features that a CNN learns when
trained and tested on different distributions of a synthetic dataset with
controlled modes of variation. We show that different loss functions lead to
different features being learned and consequently affect the generalization
ability of the classifier on unseen data. This study provides some important
insights into the design of deep learning solutions for medical imaging tasks.
Date Issued
2021-09-21
Date Acceptance
2021-08-01
Citation
2021, 12929, pp.56-64
Publisher
Springer
Start Page
56
End Page
64
Volume
12929
Copyright Statement
© 2021 Springer Nature Switzerland AG. The final publication is available at Springer via https://doi.org/10.1007/978-3-030-87444-5_6
Sponsor
Engineering & Physical Science Research Council (E
Identifier
http://arxiv.org/abs/2108.04815v1
Grant Number
RTJ13261760-1
Source
Interpretability of Machine Intelligence in Medical Image Computing at MICCAI 2021
Subjects
cs.CV
cs.CV
Notes
Accepted at iMIMIC, MICCAI 2021
Publication Status
Published
Start Date
2021-09-27
Finish Date
2021-10-01
Coverage Spatial
Strasbourg, France
Date Publish Online
2021-09-21