Key result
GINO cuts activation time map prediction error by ~43% versus GNN on synthetic LV geometries.
Why the study?
Approximately one-third of CRT recipients fail to respond partly due to suboptimal lead placement, and identifying optimal pacing sites remains challenging due to anatomical variability and limitations of individualized planning strategies.
Do geometric deep learning models accurately predict cardiac activation time maps compared to finite-element simulations in synthetic and real-world left ventricular geometries?
Do geometric deep learning models accurately predict cardiac activation time maps compared to finite-element simulations in synthetic and real-world left ventricular geometries?
Geometric deep learning models can rapidly and accurately predict left ventricular activation maps, offering a potential real-time clinical decision-support tool for personalized cardiac resynchronization therapy optimization.
Answers are grounded in this paper and cite their sources.
May enable faster LV activation modeling for CRT; leaves open prospective clinical validation.
Cardiac resynchronization therapy (CRT) is a common intervention for patients with dyssynchronous heart failure, yet approximately one-third of recipients fail to respond, partly due to suboptimal lead placement. Identifying optimal pacing sites remains challenging, largely due to patient-specific anatomical variability and limitations of current individualized planning strategies. In a step toward an in-silico approach, we develop two geometric deep learning models, based on graph neural network (GNN) and geometry-informed neural operator (GINO), to predict activation time maps on left ventricular (LV) geometries in real time. Trained on a large dataset generated from finite-element simulations spanning a wide range of synthetic LV shapes, pacing site configurations, and tissue conductivities, the GINO model outperforms the GNN on synthetic cases (1.38% vs 2.44% error), while both demonstrate comparable performance on real-world LV geometries (GINO: 4.79% vs GNN: 4.07%). Using the trained models, we develop a workflow to identify an optimal pacing site on the LV from a given activation time map and show that both models can robustly recover ground-truth subject-specific parameters from noisy inputs. In conjunction with an interactive web-based interface (https://dcsim.egr.msu.edu/), this study shows potential and motivates future extension toward a clinical decision-support tool for personalized pre-procedural CRT optimization.
No takes yet. Share an insight, caveat, or question.
Naghavi et al. (2026) studied Dyssynchronous heart failure (n=35,275). Geometry-informed neural operator (GINO) model vs. Graph neural network (GNN) model was evaluated on Prediction error of cardiac activation time maps on synthetic cases. The geometry-informed neural operator (GINO) model outperformed the graph neural network (GNN) model in predicting cardiac activation time maps on synthetic left ventricular geometries with error.
Synapse has enriched 10 related papers. Consider them for comparative context: