Dynamic memory to alleviate catastrophic forgetting in continual learning with medical imaging

Matthias Perkonigg, Johannes Hofmanninger, Christian J. Herold, James A. Brink, Oleg Pianykh, Helmut Prosch and Georg Langs ()
Additional contact information
Matthias Perkonigg: Medical University of Vienna
Johannes Hofmanninger: Medical University of Vienna
Christian J. Herold: Medical University of Vienna
James A. Brink: Massachusetts General Hospital, Harvard Medical School
Oleg Pianykh: Massachusetts General Hospital, Harvard Medical School
Helmut Prosch: Medical University of Vienna
Georg Langs: Medical University of Vienna

Nature Communications, 2021, vol. 12, issue 1, 1-12

Abstract: Abstract Medical imaging is a central part of clinical diagnosis and treatment guidance. Machine learning has increasingly gained relevance because it captures features of disease and treatment response that are relevant for therapeutic decision-making. In clinical practice, the continuous progress of image acquisition technology or diagnostic procedures, the diversity of scanners, and evolving imaging protocols hamper the utility of machine learning, as prediction accuracy on new data deteriorates, or models become outdated due to these domain shifts. We propose a continual learning approach to deal with such domain shifts occurring at unknown time points. We adapt models to emerging variations in a continuous data stream while counteracting catastrophic forgetting. A dynamic memory enables rehearsal on a subset of diverse training data to mitigate forgetting while enabling models to expand to new domains. The technique balances memory by detecting pseudo-domains, representing different style clusters within the data stream. Evaluation of two different tasks, cardiac segmentation in magnetic resonance imaging and lung nodule detection in computed tomography, demonstrate a consistent advantage of the method.

Date: 2021
References: Add references at CitEc
Citations:

Downloads: (external link)
https://www.nature.com/articles/s41467-021-25858-z Abstract (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:nat:natcom:v:12:y:2021:i:1:d:10.1038_s41467-021-25858-z

Ordering information: This journal article can be ordered from
https://www.nature.com/ncomms/

DOI: 10.1038/s41467-021-25858-z

Access Statistics for this article

Nature Communications is currently edited by Nathalie Le Bot, Enda Bergin and Fiona Gillespie

More articles in Nature Communications from Nature
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().