Overview

Neuronal networks in the brain communicate information about a subject’s intent, internal state, and external environment through electrical activity. Neurological diseases cause one or more of these networks to corrupt this communication, leading to pathological behavior and ultimately early death. Deep brain stimulation (DBS) has been used in clinical practice to suppress pathological network dynamics and restore behavior by injecting pulses of electrical current in well defined anatomical sites. Though often therapeutic, DBS comes with drawbacks: lengthy manual programming of the signal, no adaptation to patient's needs, frequent surgical battery replacements in cases where high power DBS is required, and widespread influence to nearby cognitive loops with possible adverse side effects. To address these drawbacks, we would like to design adaptive low power therapeutic DBS strategies. This requires understanding the electrophysiological dynamics in the healthy neural circuit, the diseased neural circuit, and the impact of DBS when applied to the diseased circuit.

NCSL collaborates with experts in primate and human physiology to collect electrophysiological data in neural circuits affected by Parkinson’s disease, Dystonia and Epilsepsy. The data is then used to estimate mathematical models to understand physiological markers in disease and the mechanisms of DBS. The models must be parsimonious to allow for analysis and design of closed-loop low power DBS strategies. Therefore, the goal of NCSL is to develop and apply systems-level mathematical frameworks for modeling and controlling neuronal network activity in the brain with DBS. 

Constructing a modeling framework for neural circuits is challenging because the network is distributed; fundamental units (i.e. neurons) are complex; system phenomena change nontrivially in the diseased state; network dynamics are influenced by DBS in a non-trivial way; and supporting experimental data is difficult to collect. We are developing a systems approach to construct network models that are data-driven, physiologically compliant, scalable and tractable for the design of DBS control strategies.

NCSL aims to fundamentally impact the interface between systems & control and neuroscience and introduce new opportunities for medical treatment of neurological disorders. Specific research thrusts are to (i) understand electrophysiology and patho-electrophysiology of neuronal networks and the impact of DBS control, in Parkinson’s disease and Dystonia and Epilepsy and (ii) develop new closed-loop DBS strategies to treat these diseases. Such models will eventually be developed to design new DBS strategies that are lower in power, safer and potentially more therapeutic for Parkisnon’s disease and Dystonia; and, will also be used for automatically predicting seizures in epileptic patients. Ultimately, we envision applying this methodology to other neurological disorders on the DBS horizon including Tourette Syndrome, Depression and Obsessive Compulsive Disorders, which will significantly impact the health and lifestyles of millions of patients and families worldwide.