It is difficult to imagine neuroscience in 2015 without fMRI. No single method has been more influential in studying human cognition, with colorful maps in news outlets gradually raising the public awareness of neuroscientific progress. The fact that it is possible to look inside the head of an awake, behaving person to discover not only the detailed structure of their brain but also, to some extent, what they are thinking, is truly in the sphere of science fiction. At the same time, there is only so much information conveyed with an activity map, and the trajectory of the fMRI field is presently unclear. One of the major limiting factors of this method is the fact that its substrate for estimating brain activity is in a hemodynamic blood flow-based signal, which is quite a bit removed from the electrical activity of neurons that are at the core of our cognition. While the measured fMRI signal at a given point in the brain is known to reflect local neural activity, and is in this sense better than other human methods such as EEG and MEG, it does so in a manner that is very coarse, both in its timing and in its spatial precision. There is a general need to understand how measured fMRI signals might relate to underlying neural activity. There is a range of approaches to understanding this fMRI/neural relationship. In some laboratories, experiments are aimed at carefully determining what types of neurons and molecules serve as the actual mediators for regulating blood flow. Such experiments ask, what are the biological principles, and specific mechanisms, that determine neurovascular coupling? Our approach is somewhat different and focuses on more targeted and practical questions. One such question asks how does the firing of a single neuron, or a population of neurons, contribute to the measured fMRI response within a small volume of the brain, or voxel? Consider the following: When measuring brain activity with fMRI, the smallest unit of brain activity--the electron so to speak--is a voxel that is at least a millimeter in each dimension, often much more. As neurons are much smaller and packed in high density, each voxel in the brain contain hundreds of thousands of individual neurons. We know from direct recordings of such neurons that neural responses in such a volume are diverse. One can see this local diversity when observing spontaneous brain activity, or even under conditions of visual stimulation. For example, neighboring neurons within a single voxel respond very differently when a subject is viewing a video sequence. Yet, if all of the neurons respond differently at a given moment in time, then how can the hemodynamic signal from the point in the brain be interpreted in terms of neural firing? Would the fMRI signal reflect the mean neural activity in the population? Are there particular, special neurons that determine the fMRI signal? And would the relationship remain fixed, both during periods of stimulation and rest in which visual input is minimized? At present, there are several projects investigating these and related questions in the lab, with three papers on this topic published in the last year. In one project, we are taking a multimodal approach, combining implanted multielectrodes with fMRI to directly measure the relationship between fMRI and neural signals. In one project, we are focusing on spontaneous neural activity. Specifically, we are using the spiking responses of individual neurons to create functional activity maps of the brain. Our initial results, to be presented in two posters at this years society for Neuroscience meeting (Godlove et al., SFN Abstr (2015); Mpamaugo et al, SFN Abstr (2015)) show that the firing fluctuations of individual neurons correlate with discrete and circumscribed regions in the brain. For a given neuron, these maps are repeatable across scans and across sessions. Moreover, adjacent neurons can elicit very different spatial maps, presumably reflecting their differential involvement in particular brain networks. This project is, we believe, the beginning of a research line that has the potential to yield a new perspective on the link between local neural selectivity and large-scale brain specialization. In another project studying spontaneous activity, my laboratory has recently contributed to three publications investigating spontaneous fMRI activity (Liu et al., NeuroImage (2014); Liu et al. Cerebral Cortex (2015); Monosov et al, Journal of Neuroscience (2015)). Experiments going on in the lab currently attempt to ask whether the activity in the basal forebrain may be a critical factor shaping spontaneous neural activity throughout the rest of the brain. The basal forebrain is a small region that is the origin of many long-range anatomical projections that reach virtually the entire cerebral cortex. Our initial experiments inactivating portions of this structure in animals undergoing fMRI testing suggests that the basal forebrain is strongly involved in regulation of spontaneous signals throughout the telencephalon, particularly during over transitions of arousal gauged by eye opening and closure. This work has led to the publication of two abstracts at last years Society for Neuroscience Meeting in Washington, DC (Chang et al, SFN Abstr (2014); Turchi et al, SFN Abstr (2014)), with at least one publication planned within the next year. Finally, as another approach toward understanding how the activity of neurons all within half a millimeter of one another could be so uncoupled with one another, we are presently attempting testing individual neurons with thousands of different static and dynamic visual stimuli to understand under what conditions they behave similarly and under what conditions they begin to diverge. At the same time, we are attempting to understand what aspects of this population activity most closely match the fMRI signal measured from the same point in visual cortex (Park et al., SFN Abstr (2015)). These findings may provide some insights into the complex relationship between neural responses, the relationship of neural firing within a population, and the larger relationship to hemodynamic regulation.
| Ghazizadeh, Ali; Griggs, Whitney; Leopold, David A et al. (2018) Temporal-prefrontal cortical network for discrimination of valuable objects in long-term memory. Proc Natl Acad Sci U S A 115:E2135-E2144 |
| Seidlitz, Jakob; Sponheim, Caleb; Glen, Daniel et al. (2018) A population MRI brain template and analysis tools for the macaque. Neuroimage 170:121-131 |
| Turchi, Janita; Chang, Catie; Ye, Frank Q et al. (2018) The Basal Forebrain Regulates Global Resting-State fMRI Fluctuations. Neuron 97:940-952.e4 |
| Seidlitz, Jakob; Váša, František; Shinn, Maxwell et al. (2018) Morphometric Similarity Networks Detect Microscale Cortical Organization and Predict Inter-Individual Cognitive Variation. Neuron 97:231-247.e7 |
| Liu, Xiao; de Zwart, Jacco A; Schölvinck, Marieke L et al. (2018) Subcortical evidence for a contribution of arousal to fMRI studies of brain activity. Nat Commun 9:395 |
| Takemura, Hiromasa; Pestilli, Franco; Weiner, Kevin S et al. (2017) Occipital White Matter Tracts in Human and Macaque. Cereb Cortex 27:3346-3359 |
| Park, Soo Hyun; Russ, Brian E; McMahon, David B T et al. (2017) Functional Subpopulations of Neurons in a Macaque Face Patch Revealed by Single-Unit fMRI Mapping. Neuron 95:971-981.e5 |
| Leopold, David A; Russ, Brian E (2017) Human Neurophysiology: Sampling the Perceptual World. Curr Biol 27:R71-R73 |
| Reveley, Colin; Gruslys, Audrunas; Ye, Frank Q et al. (2017) Three-Dimensional Digital Template Atlas of the Macaque Brain. Cereb Cortex 27:4463-4477 |
| Papoti, Daniel; Yen, Cecil Chern-Chyi; Hung, Chia-Chun et al. (2017) Design and implementation of embedded 8-channel receive-only arrays for whole-brain MRI and fMRI of conscious awake marmosets. Magn Reson Med 78:387-398 |
Showing the most recent 10 out of 35 publications
