Methods

Our computational methodology is built on a multi-scale approach, allowing us to investigate neurophysics from the microscopic behavior of individual ion channels up to the macroscopic dynamics of interconnected brain regions. At the cellular scale, we develop conductance-based biophysical models to simulate ion channel stochasticity, membrane noise, and synaptic input statistics. To understand the physical interface between neural tissue and applied electrical therapies, we couple electric field models (including from finite element methods) with morphologically realistic representations of arborized neurons and serpentine axonal geometries. With these in silico efforts,  we can test and optimizing electrode designs, electrical waveforms, and novel stimulation paradigms before they are evaluated in biological  systems.

Scaling up to the systems level, we utilize computational modeling to study how neural information is processed and transmitted across circuits. We construct small-scale, detailed network models to investigate dynamic phenomena such as stochastic resonance, signal fidelity, and t he rules governing memory consolidation. To capture macroscopic population dynamics, we implement multi-region, mean-field formalisms, including Wilson-Cowan architectures. These population models allow us to simulate the complex, interconnected loops of the basal ganglia, thalamus, and cortex, providing a mathematical sandbox in which to replicate the disrupted states of parkinsonism, epilepsy, dystonia, and other disorders. By bridging these computational scales, our simulation pipeline provides a rigorous, highly controllable methodology for decoding disease pathology and prototyping next-generation, closed-loop neuromodulation strategies.

 To isolate and interrogate the precise biophysical mechanisms of neural computation, we rely extensively on acute in vitro brain slice models. By preserving local microcircuitry—primarily within the hippocampus and entorhinal cortex—we can rigorously investigate how neurons, astrocytes, and microglia respond to exogenous stimulation in a highly controlled environment. Our optical methodologies center on multi-photon imaging of genetically encoded calcium indicators, allowing us to capture real-time, population-level calcium dynamics across diverse glial and neuronal networks. This high-resolution imaging is crucial for monitoring the localized, cellular-scale effects of novel neuromodulation paradigms, including temporal interference (TI) current fields and high-frequency electromagnetic fields (EMFs), without the confounding variables present in behaving animals.

In parallel with optical imaging, we employ advanced input-output slice electrophysiology to probe single-cell and synaptic biophysics. Building on the Real Time eXperimental Interface (RTXI) which we helped develop, we utilize conductance/dynamic clamp protocols to introduce simulated ion channels and artificial synaptic conductances into living neurons in real time. By precisely manipulating exogenous currents and conductances, we can systematically map the input-output transformations of individual cells. This combination of targeted electrophysiology and network-scale calcium imaging provides an essential experimental bridge, allowing us to validate our computational models and empirically test the biophysical limits of our neuroengineering tools before transitioning to in vivo systems.

To understand how pathological neural activity and restorative neuromodulation manifest at the whole-organism level, we rely extensively on in vivo models, primarily utilizing rodent systems. Our behavioral methodology pairs precise, awake-behaving electrophysiological recordings with a diverse battery of functional assays. For example, we use these models to rigorously quantify how parkinsonian states degrade complex motor behaviors, and conversely, how interventions like Deep Brain Stimulation (DBS) influence both the execution and the learning of new motor and cognitive tasks. Beyond basic locomotion, our methods extend to evaluating the effects of neuromodulation on complex social and affective behaviors. By analyzing ultrasonic vocalizations—such as the complex calls male rats make to potential mates—we can assess how subcortical stimulation modulates broad communicative and reward-based neural circuitry.

As we pioneer next-generation, minimally invasive therapies like temporal interference (TI) stimulation, our in vivo methods also encompass targeted sensory and perceptual testing, such as quantifying how focal TI current fields alter visual perception and sensory processing. To ensure these novel technologies are safe, highly selective, and clinically viable, our experimental pipeline ultimately scales up to advanced translational pre-clinical models. By evaluating our hardware and stimulation paradigms in complex mammalian nervous systems—particularly the cortico-basal ganglia-thalamo-cortical loops—we can rigorously validate device safety and optimize spatiotemporal stimulation parameters before transitioning to human clinical trials.