MixUp-MIL: Novel Data Augmentation for Multiple Instance Learning and a Study on Thyroid Cancer Diagnosis

Michael Gadermayr* (Erstautor/-in), Lukas Koller, Maximilian Tschuchnig, Lea Maria Stangassigner, Christina Erhardt-Kreutzer (Co-Autor/-in), Sébastien Couillard-Després (Co-Autor/-in), Gertie Janneke Oostingh, Anton Hittmair

*Korrespondierende/r Autor/-in für diese Arbeit

Publikation: Beitrag in Buch/Bericht/KonferenzbandKapitelBegutachtung

Abstract

Multiple instance learning is a powerful approach for whole slide image-based diagnosis in the absence of pixel- or patch-level annotations. In spite of the huge size of whole slide images, the number of individual slides is often rather small, leading to a small number of labeled samples. To improve training, we propose and investigate novel data augmentation strategies for multiple instance learning based on the idea of linear and multilinear interpolation of feature vectors within and between individual whole slide images. Based on state-of-the-art multiple instance learning architectures and two thyroid cancer data sets, an exhaustive study was conducted considering a range of common data augmentation strategies. Whereas a strategy based on to the original MixUp approach showed decreases in accuracy, a novel multilinear intra-slide interpolation method led to consistent increases in accuracy.
OriginalspracheEnglisch
TitelMedical Image Computing and Computer Assisted Intervention – MICCAI 2023
Seiten477-486
ISBN (elektronisch)978-3-031-43987-2
DOIs
PublikationsstatusVeröffentlicht - 2023
Veranstaltung26th International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI) - Vancouver, Kanada
Dauer: 8 Okt. 202312 Okt. 2023

Publikationsreihe

NameLecture Notes in Computer Science (LNCS)
ISSN (Print)0302-9743
ISSN (elektronisch)1611-3349

Konferenz mit wissenschaftlichem Inhalt

Konferenz mit wissenschaftlichem Inhalt26th International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI)
Land/GebietKanada
OrtVancouver
Zeitraum8/10/2312/10/23

Fingerprint

Untersuchen Sie die Forschungsthemen von „MixUp-MIL: Novel Data Augmentation for Multiple Instance Learning and a Study on Thyroid Cancer Diagnosis“. Zusammen bilden sie einen einzigartigen Fingerprint.

Dieses zitieren