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

Purpose: AI in radiology is hindered chiefly by: 1) Requiring large annotateddata sets. 2) Non-generalizability that limits deployment to new scanners /institutions. And 3) Inadequate explainability and interpretability. We believethat reinforcement learning can address all three shortcomings, with robust andintuitive algorithms trainable on small datasets. To the best of our knowledge,reinforcement learning has not been directly applied to computer vision tasksfor radiological images. In this proof-of-principle work, we train a deepreinforcement learning network to predict brain tumor location.  Materials and Methods: Using the BraTS brain tumor imaging database, wetrained a deep Q network on 70 post-contrast T1-weighted 2D image slices. Wedid so in concert with image exploration, with rewards and punishments designedto localize lesions. To compare with supervised deep learning, we trained akeypoint detection convolutional neural network on the same 70 images. Weapplied both approaches to a separate 30 image testing set.  Results: Reinforcement learning predictions consistently improved duringtraining, whereas those of supervised deep learning quickly diverged.Reinforcement learning predicted testing set lesion locations with 85%accuracy, compared to roughly 7% accuracy for the supervised deep network.  Conclusion: Reinforcement learning predicted lesions with high accuracy,which is unprecedented for such a small training set. We believe thatreinforcement learning can propel radiology AI well past the inherentlimitations of supervised deep learning, with more clinician-driven researchand finally toward true clinical applicability.

Deep reinforcement learning to detect brain lesions on MRI: a proof-of-concept application of reinforcement learning to medical images