CRCNS: Investigating how neural dynamics transition along the time-frequency axis across scales

While extensive evidence has shown the importance of whole-person approaches to physical and mental health, existing therapeutic interventions for neurological and mental disorders remain largely single-scale and segregated. The brain, in contrast, is a vastly multiscale and interconnected complex system, and neural dynamics transform dramatically across spatiotemporal scales, internal states, and external contexts. In particular, numerous bodies of neuroscience research have discovered methods for describing brain (dys)function in the time domain, whereas many parallel, largely disconnected efforts have discovered ways to characterize neural dynamics in the frequency domain. The goal of this project is to integrate the complementary dimensions of time and frequency into a unifying theory through the development of causal discovery methods across broad spatiotemporal scales, internal states, and external contexts. Based on extensive preliminary data, our central hypothesis (CH) is that causal mechanisms in complex networked systems such as the brain can be discovered more accurately and robustly in the frequency domain under (both of) two conditions: sufficient spatiotemporal averaging and sufficient stationarity of underlying dynamics. We test our CH through three Specific Aims, each of which combines formal mathematical analysis, causal algorithm design, large-scale numerical simulations, and in vivo electrophysiology and calcium imaging. Aim 1 tests the working hypothesis (WH) that stationary macroscopic causal mechanisms can be learned more robustly and accurately in the frequency domain. Aim 2 then tests the WH that mesoscopic causal mechanisms can best be understood using a novel form of integrated time-frequency analysis that avoids using conventional, Fourier-based decompositions, while Aim 3 tests the WH that macroscopic causal network mechanisms can be more accurately discovered using time-domain methods during the transient period following extrinsic events. The proposed research is thus highly synergistic with the NCCIH Strategic Plan, especially “Objective 1: Advance fundamental science and methods development”. Through the development of innovative causal discovery methods, the proposed research provides the computational theory needed to identify optimal intervention targets across scales and conditions. This project pioneers a transition from conventional, single-scale to a causal, systems understanding of neural information processing. Consequently, it will lay a rigorous foundation for much-needed holistic and multi-scale intervention design for neurological and mental disorders. RELEVANCE (See instructions): This project is relevant to public health because it develops an integrated framework for determining (1) where (at what nodes) and (2) how (in what “language”) to intervene for brain-related disorders. We achieve (1) by developing methods for identifying causal network mechanisms across broad spatiotemporal scales. We achieve (2) by developing a complex-systems theory to determine which analysis domain—time or frequency—is most appropriate for answering which sets of questions, at what scale, and why. Project Number: 1R01NS148491-01 | Fiscal Year: 2025 | NIH Institute/Center: National Institute of Neurological Disorders and Stroke (NINDS) | Principal Investigator: Erfan Nozari (+1 co-PI) | Institution: UNIVERSITY OF CALIFORNIA RIVERSIDE, RIVERSIDE, CA | Award Amount: $357,047 | Activity Code: R01 | Study Section: Special Emphasis Panel[ZRG1 BN-P (50)] View on NIH RePORTER: https://reporter.nih.gov/project-details/11328242