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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

While self-supervised learning (SSL) algorithms have been widely used topre-train deep models, few efforts [11] have been done to improverepresentation learning of X-ray image analysis with SSL pre-trained models. Inthis work, we study a novel self-supervised pre-training pipeline, namelyMulti-task Self-super-vised Continual Learning (MUSCLE), for multiple medicalimaging tasks, such as classification and segmentation, using X-ray imagescollected from multiple body parts, including heads, lungs, and bones.Specifically, MUSCLE aggregates X-rays collected from multiple body parts forMoCo-based representation learning, and adopts a well-designed continuallearning (CL) procedure to further pre-train the backbone subject various X-rayanalysis tasks jointly. Certain strategies for image pre-processing, learningschedules, and regularization have been used to solve data heterogeneity,overfitting, and catastrophic forgetting problems for multi-task/datasetlearning in MUSCLE.We evaluate MUSCLE using 9 real-world X-ray datasets withvarious tasks, including pneumonia classification, skeletal abnormalityclassification, lung segmentation, and tuberculosis (TB) detection. Comparisonsagainst other pre-trained models [7] confirm the proof-of-concept thatself-supervised multi-task/dataset continual pre-training could boost theperformance of X-ray image analysis.

MUSCLE: Multi-task Self-supervised Continual Learning to Pre-train Deep Models for X-ray Images of Multiple Body Parts