Edward Castillo

This study shows MRI can predict blood clot composition and contraction state using radiomic features, supporting its use for pre-interventional clot phenotyping.

This study identifies distinct metabolic signatures in stable and progressive IPF/ILD by integrating HRCT scores with metabolomic profiling.

This paper develops a scalable approach to recover recurrent dynamics in neural time series data using low-rank tensors and demonstrates its application to simulated human brain connectivity.

This paper introduces a deep learning model to analyze multiscale brain activity dynamics and demonstrates its effectiveness in capturing causal interactions across scales.