Machine Learning Technology to Automate Thoracic Aorta Dimensions by Echocardiography

Article | Jan 7, 2025
Machine Learning Technology To Automate Thoracic Aorta Dimensions By Echocardiography

This study highlights the use of a convolutional neural network (CNN) [Us2.ai] to measure key thoracic aortic parameters—LVOT, SoV, STJ, and pAA—in Aortic Stenosis (AS) patients. The CNN demonstrated comparability to cardiologist measurements on transthoracic echocardiography (TTE) and computed tomographic angiography (CTA). With its ability to deliver rapid and highly reproducible measurements, CNN technology holds promise for enhancing echocardiographic screening of aortic dilation in patients with AS.

Machine Learning Technology To Automate Thoracic Aorta Dimensions By Echocardiography
Machine Learning Technology To Automate Thoracic Aorta Dimensions By Echocardiography
Access the full publication here

Krishna, H., Dohse, C., Smith, D., Frost, M., Equilbec, C., Chin, G., Hill, M., Rodriguez Ziccardi, M. C., Slostad, B., Carter, A., Tiu, D., Darbar, D., Pellikka, P. A., & Kansal, M. (2024). Application of Machine Learning Technology to Automate Thoracic Aorta Dimensions by Echocardiography. Journal of the American Society of Echocardiography. https://doi.org/10.1016/j.echo.2024.10.017