Crowdsourced emphysema assessment

Silas Nyboe Ørting*, Veronika Cheplygina, Jens Petersen, Laura H. Thomsen, Mathilde M. W. Wille, Marleen de Bruijne

*Corresponding author for this work
2 Citations (Scopus)

Abstract

Classification of emphysema patterns is believed to be useful for improved diagnosis and prognosis of chronic obstructive pulmonary disease. Emphysema patterns can be assessed visually on lung CT scans. Visual assessment is a complex and time-consuming task performed by experts, making it unsuitable for obtaining large amounts of labeled data. We investigate if visual assessment of emphysema can be framed as an image similarity task that does not require expert. Substituting untrained annotators for experts makes it possible to label data sets much faster and at a lower cost. We use crowd annotators to gather similarity triplets and use t-distributed stochastic triplet embedding to learn an embedding. The quality of the embedding is evaluated by predicting expert assessed emphysema patterns. We find that although performance varies due to low quality triplets and randomness in the embedding, we still achieve a median F1 score of 0.58 for prediction of four patterns.

Original languageEnglish
Title of host publicationIntravascular Imaging and Computer Assisted Stenting, and Large-Scale Annotation of Biomedical Data and Expert Label Synthesis : 6th Joint International Workshops, CVII-STENT 2017 and Second International Workshop, LABELS 2017, Held in Conjunction with MICCAI 2017, Québec City, QC, Canada, September 10–14, 2017, Proceedings
EditorsM. Jorge Cardoso, Tal Arbel, Su-Lin Lee, Veronika Cheplygina, Simone Balocco, Diana Mateus, Guillaume Zahnd, Lena Maier-Hein, Stefanie Dermirci, Eric Granger, Luc Duong, Marc-André Carbonneau, Shadi Albarquoni, Gustaco Carneiro
Number of pages10
PublisherSpringer
Publication date2017
Pages126-135
ISBN (Print)978-3-319-67533-6
ISBN (Electronic)978-3-319-67534-3
DOIs
Publication statusPublished - 2017
Event2nd International Workshop on Large-Scale Annotation of Biomedical Data and Expert Label Synthesis - Québec City, Canada
Duration: 10 Sept 201714 Sept 2017
Conference number: 2

Workshop

Workshop2nd International Workshop on Large-Scale Annotation of Biomedical Data and Expert Label Synthesis
Number2
Country/TerritoryCanada
CityQuébec City
Period10/09/201714/09/2017
SeriesLecture notes in computer science
Volume10552
ISSN0302-9743

Keywords

  • Crowdsourcing
  • Emphysema
  • Similarity learning

Fingerprint

Dive into the research topics of 'Crowdsourced emphysema assessment'. Together they form a unique fingerprint.

Cite this