Fully automated cardiac function and myocardial stress analysis using deep learning

Abstract

Systems and methods for cardiac function and myocardial stress analysis include techniques and structures for classifying a set of cardiac images from a view of the set of cardiac images, detecting cardiac range and effective short-axis slice images in the set of cardiac images, determining cardiac segment locations, and determining cardiac function and myocardial stress in the set of cardiac images. Segmenting the cardiac anatomical structure for each time frame and each slice image, calculating volume-related parameters, determining a key physiological time point, calculating a myocardial transmural thickness and deriving a cardiac function measurement from the myocardial transmural thickness at the key physiological time point, estimating a dense motion field from the set of cardiac images, myocardial stresses along different myocardial directions are calculated from the dense motion field, and cardiac function measurements and myocardial stress calculations are provided to a user through a user interface.

Description

technical field [0001] The present disclosure relates generally to the study of cardiac physiology, and in particular to automating the analysis of cardiac function and myocardial stress. Background technique [0002] Cardiac function and myocardial stress analysis are critical for the diagnosis and treatment of cardiovascular disease. Cardiac function analysis usually includes measurement of cardiac chamber volume, ejection fraction, and myocardial thickness. Myocardial stress measures myocardial deformation from estimates of myocardial motion and has been shown to be a comprehensive, sensitive, and early indicator of cardiac dysfunction. These analyses are complex and require extensive domain expertise. Over the years, great efforts have been made to simplify and automate the various processes involved in cardiac function analysis and myocardial stress analysis. However, current solutions still require a lot of manual observation, manipulation of views and interpretatio...

AI Technical Summary

Problems solved by technology

However, current solutions still require extensive human observation, manipulation of views, and interpretation of results

[0003] Human intervention in these procedures typically requires extensive training and experience, often leads to problems with inter- and intra-observer variability, may provide poor analytical reproducibility, and requires additional time and effort

Images