Understanding anatomy classification through attentive response maps

Devinder Kumar, Vlado Menkovski, Graham W. Taylor, Alexander Wong

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

2 Citations (Scopus)
1 Downloads (Pure)

Abstract

One of the main challenges for broad adoption of deep learning based models such as convolutional neural networks (CNN), is the lack of understanding of their decisions. In many applications, a simpler, less capable model that can be easily understood is favorable to a black-box model that has superior performance. In this paper, we present an approach for designing CNNs based on visualization of the internal activations of the model. We visualize the model's response through attentive response maps obtained using a fractional stride convolution technique and compare the results with known imaging landmarks from the medical literature. We show that sufficiently deep and capable models can be successfully trained to use the same medical landmarks a human expert would use. Our approach allows for communicating the model decision process well, but also offers insight towards detecting biases.

Original languageEnglish
Title of host publication2018 IEEE 15th International Symposium on Biomedical Imaging, ISBI 2018
Place of PublicationPiscataway
PublisherIEEE Computer Society
Pages1130-1133
Number of pages4
ISBN (Electronic)978-1-5386-3636-7
DOIs
Publication statusPublished - 23 May 2018
Event15th IEEE International Symposium on Biomedical Imaging (ISBI 2018) - Omni Shoreham Hotel, Washington, United States
Duration: 4 Apr 20187 Apr 2018
Conference number: 15
https://biomedicalimaging.org/2018/

Conference

Conference15th IEEE International Symposium on Biomedical Imaging (ISBI 2018)
Abbreviated titleISBI18
CountryUnited States
CityWashington
Period4/04/187/04/18
Internet address

Keywords

  • Anatomy
  • CNN
  • Deep Learning
  • Visualization

Fingerprint Dive into the research topics of 'Understanding anatomy classification through attentive response maps'. Together they form a unique fingerprint.

Cite this