Scientific publications

Find Publications

May, 2023

EchoGPT, A Multimodal Language Model for Qualitative Reporting of Echocardiography

Conference: American Society of Echocardiography 35th Annual Scientific Sessions

Qualitative interpretation of echocardiography remains a nuanced and sophisticated discipline owing to the variety of cardiac pathophysiologies. Capturing and reproducing a full echocardiographic report with artificial intelligence (AI) in a manner that mimics a human echocardiographer remains the objective of AI researchers.

May, 2023

AI-Based Diagnosis of Aortic Stenosis in B-Mode Imagery

Conference: American Society of Echocardiography 35th Annual Scientific Sessions

Aortic Stenosis (AS) is a critical cardiac condition often diagnosed through echocardiography. However, the interpretation of echocardiographic data can be challenging due to variability in image quality and reader expertise.

May, 2023

Reproduction of Full Echocardiographic Report using Machine Learning from Large-scale Echocardiogram Database

Conference: American Society of Echocardiography 34th Annual Scientific Sessions

Multiple previous studies have used machine learning tools to estimate individual measurements (e.g. ejection fraction) and pathologies (e.g. aortic sclerosis) from echocardiographic images. However, to our knowledge no study has attempted to use machine learning to recreate the full echocardiographic report.

April, 2023

Automated Detection of Aortic Stenosis Using Machine Learning

Journal: Journal of American Society of Echocardiography

Aortic stenosis (AS) is a degenerative valve condition that is underdiagnosed and undertreated. Detection of AS using limited two-dimensional echocardiography could enable screening and improve appropriate referral and treatment of this condition. The aim of this study was to develop methods for automated detection of AS from limited imaging data sets. Read more ...