George Em Karniadakis

This paper develops a multiscale modeling framework to study macrophage-mediated RBC clearance in sickle cell disease and Gaucher disease, revealing altered signaling and biophysical mechanisms.

This paper develops a deep learning framework for obstructive sleep apnea classification from ECG signals, using two datasets and integrating feature engineering with transfer learning for improved generalization and personalization.

This paper introduces MR-AIV, a physics-informed AI method to reconstruct brain-wide fluid velocity fields from MRI, revealing distinct transport mechanisms in health and disease.

This paper improves and evaluates MR-AIV, a physics-informed neural network for inferring brain fluid dynamics from MRI, showing stable results across different modeling choices.

This paper develops a deep reinforcement learning algorithm for automated insulin dosing in type 1 diabetes, incorporating patient-specific carbohydrate intake and physical exercise.

This paper develops a computational platform to evaluate anti-sickling agents by integrating PK/PD models with RBC sickling kinetics, demonstrating its utility in predicting therapeutic efficacy and noncompliance outcomes for existing and potential drugs.

This study uses DPD simulations to analyze how altered RBC properties in Gaucher Disease affect blood viscosity and bone health through three shear rate domains.

This paper develops a non-invasive ML model to identify infarcted myocardium using cardiac strain data, combining synthetic rodent data with limited human data to predict LGE ground truth.

This paper develops an AI velocimetry method to infer 3D CSF flow in mice, including pressure, shear stress, and flow rates, with uncertainty quantification.

CERA is a Shiny web tool that uses ChatGPT to simplify ROC curve analysis and cutoff determination for biomarker studies.

This study uses computational models to show that deformable interendothelial slits in the spleen better predict RBC filtration compared to rigid models, improving understanding of splenic hemodynamics and filtration mechanisms.

The paper develops a deep learning algorithm that integrates ordinary differential equations into neural networks to infer hidden dynamics and parameters from sparse experimental data.