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

Left-ventricular ejection fraction (LVEF) is an important indicator of heartfailure. Existing methods for LVEF estimation from video require large amountsof annotated data to achieve high performance, e.g. using 10,030 labeledechocardiogram videos to achieve mean absolute error (MAE) of 4.10. Labelingthese videos is time-consuming however and limits potential downstreamapplications to other heart diseases. This paper presents the firstsemi-supervised approach for LVEF prediction. Unlike general video predictiontasks, LVEF prediction is specifically related to changes in the left ventricle(LV) in echocardiogram videos. By incorporating knowledge learned frompredicting LV segmentations into LVEF regression, we can provide additionalcontext to the model for better predictions. To this end, we propose a novelCyclical Self-Supervision (CSS) method for learning video-based LVsegmentation, which is motivated by the observation that the heartbeat is acyclical process with temporal repetition. Prediction masks from oursegmentation model can then be used as additional input for LVEF regression toprovide spatial context for the LV region. We also introduce teacher-studentdistillation to distill the information from LV segmentation masks into anend-to-end LVEF regression model that only requires video inputs. Results showour method outperforms alternative semi-supervised methods and can achieve MAEof 4.17, which is competitive with state-of-the-art supervised performance,using half the number of labels. Validation on an external dataset also showsimproved generalization ability from using our method. Our code is available athttps://github.com/xmed-lab/CSS-SemiVideo.

Cyclical Self-Supervision for Semi-Supervised Ejection Fraction Prediction from Echocardiogram Videos