Key result
A graph neural network using myocardium and fibrosis segmentations from cardiac MRI predicted ICD therapy with an AUC of 0.600, outperforming baseline models based on LVEF and LGE.
Why the study?
Many patients receiving an ICD for high risk of ventricular arrhythmias and sudden cardiac death never receive appropriate therapy, creating a need for automatic fibrosis segmentation and better risk stratification.
Does a deep learning network using LGE CMR fibrosis segmentations improve the prediction of appropriate device therapy in ICD patients compared to current guideline markers?
Does a deep learning network using LGE CMR fibrosis segmentations improve the prediction of appropriate device therapy in ICD patients compared to current guideline markers?
Answers are grounded in this paper and cite their sources.
A novel deep learning approach using LGE CMR fibrosis segmentation shows promise in predicting appropriate ICD therapy, outperforming current guideline-based markers.
Many patients at high risk of life-threatening ventricular arrhythmias (VA) and sudden cardiac death (SCD) who received an implantable cardioverter defibrillator (ICD), never receive appropriate device therapy. The presence of fibrosis on LGE CMR imaging isshown to be associated with increased risk of VA.Therefore, there is a strong need for both automatic segmentation and quantification of cardiac fibrosis as well as better risk stratification for SCD.This study first presents a novel two-stage deep learning network for the segmentation of left ventricle myocardium and fibrosis on LGE CMR images.Secondly it aims to effectively predict device therapy in ICD patients by using a graph neural network approach which incorporates both myocardium and fibrosis features as well as the left ventricle geometry.Our segmentation network outperforms previous stateof-the-art methods on 2D CMR data, reaching a Dice score of 0.82 and 0.77 on myocardium and fibrosis segmentation, respectively.The ICD therapy prediction network reaches an AUC of 0.60 while using only CMR data and outperforms baseline methods based on current guideline markers for ICD implantation.This work lays a strong basis for future research on improved risk stratification for VA and SCD.
No takes yet. Share an insight, caveat, or question.
Lieshout et al. (2022) studied Patients with an implantable cardioverter defibrillator (ICD) (n=515). Graph neural network (myo-GNN) using myocardium and fibrosis segmentations vs. Logistic regression based on LVEF or LGE was evaluated on Prediction of ICD device therapy (AUC). A graph neural network using myocardium and fibrosis segmentations from cardiac MRI predicted ICD therapy with an AUC of 0.600, outperforming baseline models based on LVEF and LGE.
Synapse has enriched 10 related papers. Consider them for comparative context: