Artificial Intelligence-Enhanced Analysis of Echocardiography-Based Radiomic Features for Myocardial Hypertrophy Detection and Etiology Differentiation

doi:10.1161/circimaging.124.017436

Artificial Intelligence-Enhanced Analysis of Echocardiography-Based Radiomic Features for Myocardial Hypertrophy Detection and Etiology Differentiation

Circ Cardiovasc ImagingPublished 27 Mar 2025development of an artificial intelligence algorithm using echocardiography-based radiomics

Read the original article ↗

Cite

Authors' conclusion

This study confirms that artificial intelligence-enhanced echocardiography-based radiomics effectively differentiate the etiology of LVH, highlighting the potential of artificial intelligence-driven texture and geographic analysis in LVH evaluation.

Study context

Population: Subjects assessed for left ventricular hypertrophy (LVH) and its etiologies: hypertrophic cardiomyopathy (HCM), cardiac amyloidosis (CA) and hypertensive heart disease (HHD)developmental data sets 867 · independent external test set 619

Comparison: radiomics-based LightGBM model compared with the logistic regression model using conventional echocardiographic parameters

Methodological detailsmultiple medical centers (developmental data sets), independent external test set from a single tertiary medical center, 4 echocardiographic views, Shapley Additive Explanations analysis

Multiple medical centers (developmental data sets)
Independent external test set from a single tertiary medical center
4 echocardiographic views
Shapley Additive Explanations analysis

Main finding

As reported in the abstract

Area under the curve, external test set

Group	HCM	CA	HHD
Radiomics-based LightGBM model	0.96	0.89	0.86

Radiomics-based LightGBM model

HCM0.96

CA0.89

HHD0.86

Further findings

Outcome	Radiomics-based LightGBM model	Logistic regression model using conventional echocardiographic parameters
Sensitivity
HCM	0.89	0.80
CA	0.80	0.80
HHD	0.75	0.33
F1-score
HCM	0.87	0.57
CA	0.84	0.72
HHD	0.82	0.50

Feature analysis

Harmonization-driven textures played a key role in differentiating HCM
Conventional textures and myocardial thickness were influential in differentiating CA and HHD

Original publication

Artificial Intelligence-Enhanced Analysis of Echocardiography-Based Radiomic Features for Myocardial Hypertrophy Detection and Etiology Differentiation. Circ Cardiovasc Imaging. Published 27 March 2025. doi:10.1161/circimaging.124.017436

Read the original article ↗PubMed

Summarised from the published abstract; every number is checked against it