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

Unsupervised anomaly detection (UAD) aims to find anomalous images byoptimising a detector using a training set that contains only normal images.UAD approaches can be based on reconstruction methods, self-supervisedapproaches, and Imagenet pre-trained models. Reconstruction methods, whichdetect anomalies from image reconstruction errors, are advantageous becausethey do not rely on the design of problem-specific pretext tasks needed byself-supervised approaches, and on the unreliable translation of modelspre-trained from non-medical datasets. However, reconstruction methods may failbecause they can have low reconstruction errors even for anomalous images. Inthis paper, we introduce a new reconstruction-based UAD approach that addressesthis low-reconstruction error issue for anomalous images. Our UAD approach, thememory-augmented multi-level cross-attentional masked autoencoder (MemMC-MAE),is a transformer-based approach, consisting of a novel memory-augmentedself-attention operator for the encoder and a new multi-level cross-attentionoperator for the decoder. MemMCMAE masks large parts of the input image duringits reconstruction, reducing the risk that it will produce low reconstructionerrors because anomalies are likely to be masked and cannot be reconstructed.However, when the anomaly is not masked, then the normal patterns stored in theencoder's memory combined with the decoder's multi-level cross attention willconstrain the accurate reconstruction of the anomaly. We show that our methodachieves SOTA anomaly detection and localisation on colonoscopy, pneumonia, andcovid-19 chest x-ray datasets.

Unsupervised Anomaly Detection in Medical Images with a Memory-augmented Multi-level Cross-attentional Masked Autoencoder