Contrastive Learning for View Classification of Echocardiograms

Agisilaos Chartsias*, Shan Gao, Angela Mumith, Jorge Oliveira, Kanwal Bhatia, Bernhard Kainz, Arian Beqiri

*Corresponding author for this work

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

9 Citations (Scopus)

Abstract

Analysis of cardiac ultrasound images is commonly performed in routine clinical practice for quantification of cardiac function. Its increasing automation frequently employs deep learning networks that are trained to predict disease or detect image features. However, such models are extremely data-hungry and training requires labelling of many thousands of images by experienced clinicians. Here we propose the use of contrastive learning to mitigate the labelling bottleneck. We train view classification models for imbalanced cardiac ultrasound datasets and show improved performance for views/classes for which minimal labelled data is available. Compared to a naïve baseline model, we achieve an improvement in F1 score of up to 26% in those views while maintaining state-of-the-art performance for the views with sufficiently many labelled training observations.

Original languageEnglish
Title of host publicationSimplifying Medical Ultrasound - Second International Workshop, ASMUS 2021, Held in Conjunction with MICCAI 2021, Proceedings
EditorsJ. Alison Noble, Stephen Aylward, Alexander Grimwood, Zhe Min, Su-Lin Lee, Yipeng Hu
PublisherSpringer Science and Business Media Deutschland GmbH
Pages149-158
Number of pages10
ISBN (Print)9783030875824
DOIs
Publication statusPublished - 2021
Event2nd International Workshop on Advances in Simplifying Medical UltraSound, ASMUS 2021 held in conjunction with 24th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2021 - Strasbourg, France
Duration: 27 Sept 202127 Sept 2021

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12967 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference2nd International Workshop on Advances in Simplifying Medical UltraSound, ASMUS 2021 held in conjunction with 24th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2021
Country/TerritoryFrance
CityStrasbourg
Period27/09/202127/09/2021

Keywords

  • Classification
  • Contrastive learning
  • Echocardiography

Fingerprint

Dive into the research topics of 'Contrastive Learning for View Classification of Echocardiograms'. Together they form a unique fingerprint.

Cite this