New self-supervised method accurately detects cardiac phases in echocardiography

Researchers have developed a novel self-supervised method for detecting key cardiac phases in echocardiography, specifically end-diastole (ED) and end-systole (ES). This new approach constrains the latent motion component of the cardiac cycle to a single-parameter latent orbit, effectively creating a global linear trajectory in latent space. This allows for direct identification of ED and ES from the learned phase signal, offering a more interpretable representation of the cardiac cycle while still accommodating irregular heartbeats. Trained on the EchoNet-Dynamic dataset without annotations, the model demonstrates improved performance in ED localization and matches existing methods for ES localization, using a more constrained representation and fewer training epochs. AI

IMPACT This method could lead to more objective and efficient cardiac function quantification, reducing inter-operator variability in clinical settings.

RANK_REASON The cluster contains an academic paper detailing a new method for a specific domain. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• arXiv
• cs.CV
• EchoNet-Dynamic

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →