The blood oxygenation level dependent (BOLD) magnetic resonance imaging (MRI) fluctuations used to map functional connectivity contain a wealth of information about neural activity and physiological processes in the brain. Most functional connectivity studies wish to detect time-varying activity related to cognition and information processing, and view the presence of other contributors to the spontaneous BOLD fluctuations as a complication. However, evidence is growing that sources of ?noise? in the BOLD signal contain clinically- relevant information about activity at different spatial and temporal scales. The challenge lies in separating contributions from different processes so that selective sensitivity to the process of interest can be achieved. We propose to combine spatial, spectral and temporal signal characteristics with multi-modal imaging to separate the BOLD fluctuations into four components with different spatial and temporal scales: 1) a quasiperiodic spatiotemporal pattern (QPP) linked to infraslow electrical activity; 2) oscillations that arise from properties of the vasculature; 3) global signal variations that do not reflect local neural processing; and 4) the remaining variability, which should have increased sensitivity to time-varying interactions between regions. The two key elements that make the isolation of BOLD components possible are the direct measurement of neural activity in conjunction with imaging experiments in the rat model, and dynamic analysis techniques that can capture spatial and temporal patterns in the imaging and recording data. While the foundational work described in this proposal will be performed in the rat, the tools we develop will be optimized and applied to standard resting state functional MRI (rs-fMRI) studies in humans. Our preliminary data shows that the BOLD signal contains contributions from two separable types of neural activity: infraslow activity, which produces quasiperiodic spatiotemporal patterns of BOLD activation; and activity in typical EEG bands, which is more closely tied to time-varying activity between areas. Using only analytical tools, we show that we can separate and identify similar processes in human data, a strong argument for the ultimate translatability of these techniques. We also show that the QPPs alone account for the differences in connectivity observed between patients with major depressive disorder and healthy controls, which demonstrates how selective analysis methods can aid in the diagnosis of psychiatric and neurological disorders and provide new insight into the alterations in connectivity that many disorders exhibit. We exhibit preliminary evidence for both neural and vascular contributions to the global BOLD signal, and describe a method for mapping the contribution of vascular oscillations.
Specific aims are: 1.Determine the neural and hemodynamic correlates of the global BOLD signal; 2. Characterize the contributions of vascular oscillations; 3.Distinguish bandlimited contributions from BOLD correlates of 1/f? activity; 4. Translate findings to human studies.

Public Health Relevance

We have developed a new method of analysis that isolates four components of the functional MRI signal that coexist on different spatial and temporal scales. Our working hypothesis is that each component may be affected in a different way by neurological and psychiatric disorders, so that the isolation of these components may improve the diagnosis and evaluation of brain dysfunction. In addition, removal of non-neural components can improve sensitivity to neural activity for clinical research and basic neuroscience. In this proposal, we characterize and validate the contribution of each component using multimodal imaging in the rat, then optimize the method for translation to human studies.

National Institute of Health (NIH)
National Institute of Mental Health (NIMH)
Research Project (R01)
Project #
Application #
Study Section
Special Emphasis Panel (ZMH1)
Program Officer
Kim, Douglas Sun-IL
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
Emory University
Biomedical Engineering
Schools of Medicine
United States
Zip Code
Pan, Wen-Ju; Lee, Seung Yup; Billings, Jacob et al. (2018) Detection of neural light-scattering activity in vivo: optical transmittance studies in the rat brain. Neuroimage 179:207-214
Belloy, Michaël E; Shah, Disha; Abbas, Anzar et al. (2018) Quasi-Periodic Patterns of Neural Activity improve Classification of Alzheimer's Disease in Mice. Sci Rep 8:10024
Yousefi, Behnaz; Shin, Jaemin; Schumacher, Eric H et al. (2018) Quasi-periodic patterns of intrinsic brain activity in individuals and their relationship to global signal. Neuroimage 167:297-308
Belloy, Michaël E; Naeyaert, Maarten; Abbas, Anzar et al. (2018) Dynamic resting state fMRI analysis in mice reveals a set of Quasi-Periodic Patterns and illustrates their relationship with the global signal. Neuroimage 180:463-484
Billings, Jacob; Keilholz, Shella (2018) The Not-So-Global Blood Oxygen Level-Dependent Signal. Brain Connect 8:121-128
Billings, Jacob C W; Medda, Alessio; Shakil, Sadia et al. (2017) Instantaneous brain dynamics mapped to a continuous state space. Neuroimage 162:344-352
Keilholz, Shella D; Pan, Wen-Ju; Billings, Jacob et al. (2017) Noise and non-neuronal contributions to the BOLD signal: applications to and insights from animal studies. Neuroimage 154:267-281