Multiple Instance Learning for Heterogeneous Images: Training a CNN for Histopathology

By Heather D. Couture, James Stephen Marron, Charles M. Perou, M. Troester, Marc Niethammer. 2018

69 citations recorded by the source corpus.

Multiple instance (MI) learning with a convolutional neural network enables end-to-end training in the presence of weak image-level labels. We propose a new method for aggregating predictions from smaller regions of the image into an image-level classification by using the quantile function. The quantile function provides a more complete description of the heterogeneity within each image, improving image-level classification. We also adapt image augmentation to the MI framework by randomly selecting cropped regions on which to apply MI aggregation during each epoch of training. This provides a mechanism to study the importance of MI learning. We validate our method on five different classification tasks for breast tumor histology and provide a visualization method for interpreting local image classifications that could lead to future insights into tumor heterogeneity.

Subfields: Artificial Intelligence, Biophysics, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Biophysics

In Computer Vision and Pattern Recognition

Citations in the corpus

A Survey on Explainable Artificial Intelligence (XAI): Toward Medical XAI (2019, methodology): “In the medical field (see later section), [34] [55] [56] [57] [58] have studied methods employing visual explanations.”
Unbox the black-box for the medical explainable AI via multi-modal and multi-centre data fusion: A mini-review, two showcases and beyond (2021, methodology): “Coupling with CNN, Couture et al. [85] proposed a multi-instance aggregation approach to classify breast tumour tissue microarray for various clinical tasks, e.g., histologic subtype classification, and the derived super-pixel maps could highlight the area where the tumour cells were and each mark…”
Data-efficient and weakly supervised computational pathology on whole-slide images (2020, methodology): “In contrast, CLAM is applicable to general multi-class classiﬁcation as instead of using max pooling or other aggregation functions ( e.g. mean operator, generalized mean, log-sum-exp, the quantile function, Noisy-OR, Nosiy-And 35–37 , etc. ) of limited ﬂexibility for dataset or problem-speciﬁc tuni”
Artificial intelligence in histopathology: enhancing cancer research and clinical oncology (2022, background): “As early as 2010, several groups experimented with MIL in pathology image analysis 112,113 , but it took several years for these ideas to be more broadly used 114 .”