Quantitative imaging analysis in hypertrophic cardiomyopathy: phenotypic differentiation to prognostic stratification

Hypertrophic cardiomyopathy (HCM) is the most common inherited cardiac disorder and remains associated with sudden cardiac death (SCD), heart failure, and diagnostic uncertainty due to overlap with other causes of left ventricular hypertrophy, including hypertensive heart disease and cardiac amyloidosis. This review aimed to evaluate the evolving applications of radiomics in the context of related quantitative imaging techniques. A review was conducted of studies applying radiomics, texture analysis, machine learning, and deep learning to cardiac imaging in HCM. The evidence base derives predominantly from CMR, with a smaller and more recent body of work in echocardiography and CT. Current evidence suggests that radiomics can extract quantitative features of myocardial texture, shape, and signal heterogeneity beyond visual assessment. These approaches have shown promise in differentiating HCM from phenocopies, identifying myocardial fibrosis without contrast administration, and improving the prediction of adverse outcomes, including SCD and heart failure. Several studies reported incremental value over conventional imaging markers such as wall thickness, global T1 values, and binary late gadolinium enhancement. However, the literature remains limited by retrospective study design, small cohorts, heterogeneity in imaging acquisition and segmentation, and insufficient external validation. Radiomics has emerged as a promising adjunct to conventional cardiac imaging in HCM, with the potential to be adapted into a clinical decision support tool. Its future clinical role will depend on methodological standardization and robust external validation.