Key result
Automated machine learning accurately quantifies CMR LGE scar in HCM, matching experts with ~0.92 correlation.
Why the study?
CMR LGE scar quantification is important for risk stratifying HCM patients, motivating the development of an automated ML model to contour LV borders and quantify scar.
Does a 2-dimensional convolutional neural network accurately and automatically quantify left ventricular scar on CMR LGE images in patients with hypertrophic cardiomyopathy compared to expert manual segmentation?
Observational (n=307)
Yes
Does a 2-dimensional convolutional neural network accurately and automatically quantify left ventricular scar on CMR LGE images in patients with hypertrophic cardiomyopathy compared to expert manual segmentation?
Effect estimate: r = 0.92
A fully automated machine learning algorithm can rapidly and accurately quantify left ventricular scar burden from CMR LGE images in HCM patients without the need for manual image pre-processing.
May enable rapid automated scar quantification in HCM CMR; leaves open prospective validation before clinical adoption.
Scar quantification on cardiovascular magnetic resonance (CMR) late gadolinium enhancement (LGE) images is important in risk stratifying patients with hypertrophic cardiomyopathy (HCM) due to the importance of scar burden in predicting clinical outcomes. We aimed to develop a machine learning (ML) model that contours left ventricular (LV) endo- and epicardial borders and quantifies CMR LGE images from HCM patients.We retrospectively studied 2557 unprocessed images from 307 HCM patients followed at the University Health Network (Canada) and Tufts Medical Center (USA). LGE images were manually segmented by two experts using two different software packages. Using 6SD LGE intensity cutoff as the gold standard, a 2-dimensional convolutional neural network (CNN) was trained on 80% and tested on the remaining 20% of the data. Model performance was evaluated using the Dice Similarity Coefficient (DSC), Bland-Altman, and Pearson's correlation. The 6SD model DSC scores were good to excellent at 0.91 ± 0.04, 0.83 ± 0.03, and 0.64 ± 0.09 for the LV endocardium, epicardium, and scar segmentation, respectively. The bias and limits of agreement for the percentage of LGE to LV mass were low (-0.53 ± 2.71%), and correlation high (r = 0.92). This fully automated interpretable ML algorithm allows rapid and accurate scar quantification from CMR LGE images. This program does not require manual image pre-processing, and was trained with multiple experts and software, increasing its generalizability.
No takes yet. Share an insight, caveat, or question.
Navidi et al. (2023) conducted an observational in Hypertrophic cardiomyopathy (HCM) (n=307). 2-dimensional convolutional neural network (CNN) machine learning model vs. Manual segmentation by experts (6SD LGE intensity cutoff) was evaluated on Model performance evaluated using Dice Similarity Coefficient (DSC), Bland-Altman, and Pearson's correlation (r = 0.92). A fully automated machine learning algorithm accurately quantified scar from CMR LGE images in HCM patients, achieving a high correlation (r = 0.92) with manual expert segmentation.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: