English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

One issue with computer based histopathology image analysis is that the sizeof the raw image is usually very large. Taking the raw image as input to thedeep learning model would be computationally expensive while resizing the rawimage to low resolution would incur information loss. In this paper, we presenta novel deep hybrid attention approach to breast cancer classification. Itfirst adaptively selects a sequence of coarse regions from the raw image by ahard visual attention algorithm, and then for each such region it is able toinvestigate the abnormal parts based on a soft-attention mechanism. A recurrentnetwork is then built to make decisions to classify the image region and alsoto predict the location of the image region to be investigated at the next timestep. As the region selection process is non-differentiable, we optimize thewhole network through a reinforcement approach to learn an optimal policy toclassify the regions. Based on this novel Look, Investigate and Classifyapproach, we only need to process a fraction of the pixels in the raw imageresulting in significant saving in computational resources without sacrificingperformances. Our approach is evaluated on a public breast cancerhistopathology database, where it demonstrates superior performance to thestate-of-the-art deep learning approaches, achieving around 96\% classificationaccuracy while only 15% of raw pixels are used.

Look, Investigate, and Classify: A Deep Hybrid Attention Method for Breast Cancer Classification