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

Transfer learning from natural image to medical image has established as oneof the most practical paradigms in deep learning for medical image analysis.However, to fit this paradigm, 3D imaging tasks in the most prominent imagingmodalities (e.g., CT and MRI) have to be reformulated and solved in 2D, losingrich 3D anatomical information and inevitably compromising the performance. Toovercome this limitation, we have built a set of models, called GenericAutodidactic Models, nicknamed Models Genesis, because they are created exnihilo (with no manual labeling), self-taught (learned by self-supervision),and generic (served as source models for generating application-specific targetmodels). Our extensive experiments demonstrate that our Models Genesissignificantly outperform learning from scratch in all five target 3Dapplications covering both segmentation and classification. More importantly,learning a model from scratch simply in 3D may not necessarily yieldperformance better than transfer learning from ImageNet in 2D, but our ModelsGenesis consistently top any 2D approaches including fine-tuning the modelspre-trained from ImageNet as well as fine-tuning the 2D versions of our ModelsGenesis, confirming the importance of 3D anatomical information andsignificance of our Models Genesis for 3D medical imaging. This performance isattributed to our unified self-supervised learning framework, built on a simpleyet powerful observation: the sophisticated yet recurrent anatomy in medicalimages can serve as strong supervision signals for deep models to learn commonanatomical representation automatically via self-supervision. As open science,all pre-trained Models Genesis are available athttps://github.com/MrGiovanni/ModelsGenesis.

Models Genesis: Generic Autodidactic Models for 3D Medical Image Analysis