Developing biophysics-inspired, -informed, and -constrained artificial intelligence. Leveraging spiking neural network architectures and biophysical neuron models to create brain-inspired computing systems and AI models grounded in biological realism.
Building biophysically detailed neuron models constrained by multimodal data sets (patch-clamp electrophysiology, morphological reconstructions, single-cell transcriptomics, etc.). Using evolutionary and gradient-based optimization to generate models across cell types to create digital twins of biological neurons.
Bottom-up data-driven models of human neural circuits, integrating multimodal data to understand brain function and dysfunction. Simulating brain circuits with tens of thousands of bio-realistic and interconnected neurons to understand computational principles of the brain and the biophysics behind them.
Recording from living human cortical slices obtained during therapeutic neurosurgery at Cedars-Sinai. Simultaneous intracellular and extracellular recordings from identified neurons reveal how single cells and cell types organize human cortical circuit activity, with bio-realistic human circuit models guiding and interpreting the experiments.
Investigating cellular and circuit-level mechanisms underlying neurological diseases including epilepsy and glioma. Using patient-derived human brain tissue and multi-modal characterization to build computational models of seizure dynamics and aberrant neural activity.
Developing and applying computational methods to enhance monitoring capabilities and refine modulation strategies to control brain activity, including high-density neural probes and deep brain stimulation. Investigating how extracellular electric fields interact with neural circuits at the cellular and cell type level.