Radiological Reports Improve Pre-training for Localized Imaging Tasks on Chest X-Rays

Philip Müller, Georgios Kaissis, Congyu Zou, Daniel Rueckert

Publikation: Beitrag in Buch/Bericht/KonferenzbandKonferenzbeitragBegutachtung

7 Zitate (Scopus)

Abstract

Self-supervised pre-training on unlabeled images has shown promising results in the medical domain. Recently, methods using text-supervision from companion text like radiological reports improved upon these results even further. However, most works in the medical domain focus on image classification downstream tasks and do not study more localized tasks like semantic segmentation or object detection. We therefore propose a novel evaluation framework consisting of 18 localized tasks, including semantic segmentation and object detection, on five public chest radiography datasets. Using our proposed evaluation framework, we study the effectiveness of existing text-supervised methods and compare them with image-only self-supervised methods and transfer from classification in more than 1200 evaluation runs. Our experiments show that text-supervised methods outperform all other methods on 13 out of 18 tasks making them the preferred method. In conclusion, image-only contrastive methods provide a strong baseline if no reports are available while transfer from classification, even in-domain, does not perform well in pre-training for localized tasks.

OriginalspracheEnglisch
TitelMedical Image Computing and Computer Assisted Intervention – MICCAI 2022 - 25th International Conference, Proceedings
Redakteure/-innenLinwei Wang, Qi Dou, P. Thomas Fletcher, Stefanie Speidel, Shuo Li
Herausgeber (Verlag)Springer Science and Business Media Deutschland GmbH
Seiten647-657
Seitenumfang11
ISBN (Print)9783031164422
DOIs
PublikationsstatusVeröffentlicht - 2022
Veranstaltung25th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2022 - Singapore, Singapur
Dauer: 18 Sept. 202222 Sept. 2022

Publikationsreihe

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Band13435 LNCS
ISSN (Print)0302-9743
ISSN (elektronisch)1611-3349

Konferenz

Konferenz25th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2022
Land/GebietSingapur
OrtSingapore
Zeitraum18/09/2222/09/22

Fingerprint

Untersuchen Sie die Forschungsthemen von „Radiological Reports Improve Pre-training for Localized Imaging Tasks on Chest X-Rays“. Zusammen bilden sie einen einzigartigen Fingerprint.

Dieses zitieren