AI-based quantitative assessment of LGE is an independent predictor of adverse cardiovascular outcomes in patients with HCM and may represent a clinically meaningful imaging biomarker for risk stratification.
The high LGE group (≥15%) experienced a significantly higher incidence of the primary composite outcome compared with the low LGE group.
High AI-quantified LGE burden (≥15%) was independently associated with the primary outcome after adjustment for confounders.
ROC analysis identified an optimal LGE cutoff of 8% for predicting the primary outcome.
Patients with LGE ≥8% had markedly increased rates of both the primary outcome and SCD/SCD-equivalent events compared to those with LGE <8%.
The study used an AI-based segmentation algorithm with the 6-standard deviation method to quantify LGE on cardiovascular magnetic resonance imaging.
This research suggests that using artificial intelligence to measure the amount of scar tissue in the heart—detected through a specialized MRI technique called late gadolinium enhancement (LGE)—can help predict dangerous heart events in people with hypertrophic cardiomyopathy (HCM), a condition where the heart muscle becomes abnormally thickened. In a group of 142 HCM patients followed for about five years, those with high levels of heart scarring (15% or more of heart tissue affected) were more than four times as likely to experience serious cardiovascular events, including sudden cardiac death, compared to those with lower scarring levels. The study also found that an even lower threshold of 8% scar tissue, identified by analyzing the data mathematically, was a strong dividing line: patients with 8% or more scarring had an 18.2% rate of serious cardiac events, while patients below that threshold had zero events during follow-up. The AI system analyzed MRI images to precisely measure scar burden, offering a potentially more objective and consistent approach than traditional manual methods. This research suggests that AI-assisted measurement of heart scarring from MRI scans could be a valuable tool for identifying HCM patients who are at higher risk for life-threatening heart events, potentially helping doctors make better-informed decisions about monitoring and treatment strategies such as implantable defibrillators. The findings point to specific scar tissue thresholds (8% and 15%) that may be clinically meaningful benchmarks, though further research in larger and more diverse populations would be needed to confirm these cutoffs.
Jeong Y, Park J, Choi K, Nam J, Lee C, Son J, et al.. (2026). Prognostic significance of artificial intelligence-quantified late gadolinium enhancement on cardiac magnetic resonance in hypertrophic cardiomyopathy.. The Korean journal of internal medicine. https://doi.org/10.3904/kjim.2026.065