Multiple Instance Learning for Heterogeneous Images: Training a CNN for Histopathology
This work addresses classification challenges in medical imaging for histopathology, offering incremental improvements in handling heterogeneous images.
The authors tackled the problem of classifying histopathology images with weak labels by proposing a multiple instance learning method using the quantile function for aggregation and adapted image augmentation, achieving improved image-level classification validated on five breast tumor histology tasks.
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.