What is feasible with imaging human brain function and connectivity using functional magnetic resonance imaging

Abstract

When we consider all of the methods we employ to detect brain function, from electrophysiology to optical techniques to functional magnetic resonance imaging (fMRI), we do not really have a 'golden technique' that meets all of the needs for studying the brain. We have methods, each of which has significant limitations but provide often complimentary information. Clearly, there are many questions that need to be answered about fMRI, which unlike other methods, allows us to study the human brain. However, there are also extraordinary accomplishments or demonstration of the feasibility of reaching new and previously unexpected scales of function in the human brain. This article reviews some of the work we have pursued, often with extensive collaborations with other co-workers, towards understanding the underlying mechanisms of the methodology, defining its limitations, and developing solutions to advance it. No doubt, our knowledge of human brain function has vastly expanded since the introduction of fMRI. However, methods and instrumentation in this dynamic field have evolved to a state that discoveries about the human brain based on fMRI principles, together with information garnered at a much finer spatial and temporal scale through other methods, are poised to significantly accelerate in the next decade.This article is part of the themed issue 'Interpreting BOLD: a dialogue between cognitive and cellular neuroscience'.

Keywords: brain; cortical column; cortical layers; functional imaging; neurovascular; resting state.

Figure 1.

Single condition functional mapping of orientation columns in the cat brain using detection of cerebral blood flow (CBF) with MRI. Relative blood flow changes in response to two orthogonal grating orientations (45° and 135°) are shown. In the relative blood flow versus time plot, the blood flow changes in ‘activated’ versus other regions are displayed in solid red and dashed grey/black curves, respectively for the 45° stimulus, and vise versa for the 135° stimulus. Colour image shows an image obtained by thresholding based on blood flow change to suppress the regions showing the lower response for one orientation. Adapted from [6].

Figure 2.

As in figure 1, but images obtained with gradient echo BOLD fMRI for four different orientations of the stimulus.

Figure 3.

Extravascular relaxation rate changes for R₂ and R₂* (equal to 1/T₂, and 1/T₂*, respectively) induced by simulated blood vessels with a magnetic susceptibility difference between blood vessel interior and exterior (basis of extravascular BOLD effect), shown as a function of blood vessel radius and magnetic susceptibility induced different frequency shifts (in Hz) across the blood vessel. The numbers 32, 48 and 64 Hz correspond to increasing magnetic field strength B₀ at a constant deoxyhaemoglobin concentration (approx. 3, 5, and 7 T at physiological venous deoxyhemoglobin levels) or increasing deoxyhaemoglobin concentration at a constant B₀. GE, gradient echo; SE, spin echo. From Uludag & Ugurbil [28].

Figure 4.

Simulation of spin echo (SE) detected fractional signal changes (ΔS/S) induced by neuronal activity at TE = tissue T₂. Physiological parameters used: for microvasculature: Microvascular CBV = 2.5%, composed of 20% arteriole (d = 16 µ), 40% capillary (d = 5 µ), and 40% venule (d = 16 µ). ΔCBV is taken as either as 0 or 16% in all. For macrovasculature vessels: CBV = 5%; diameter, d = 200 µ; 90° to B_o (worst case scenario); ΔCBV = 0; a larger ΔCBV would reduce the contribution coming from this component. Adapted from Uludag et al. [14].

Figure 5.

Functional contrast-to-noise ratio (fCNR) for SE BOLD fMRI will be proportional to the curves shown, which are obtained by multiplying ΔS/S in figure 4 with , normalized to the value at 1.5 T. These plots are valid only in the limit the noise in the fMRI time series is dominated by thermal noise of the image, which is the case, for example, for high-resolution imaging at the level of cortical columns and layers. Plots generated by Uludağ K using data from Uludağ et al. [14].

Figure 6.

Functional contrast-to-noise ratio (fCNR) for GE BOLD fMRI. As in figure 5 but calculated using ΔS/S for GE BOLD fMRI data from Uludağ et al. [14]. Microvascular contribution is calculated with ΔCBV = 0% (dashed line) or 16% (solid line), as in figure 5.

Figure 7.

Functional maps of orientation and ocular dominance columns in the human brain. Obtained at 7 T using SE fMRI. Adopted from Yacoub et al. [56].

Figure 8.

Ocular dominance column functional images obtained either by spin echo (SE) or gradient echo (GE) fMRI on two separate occasions on the same individual. Each voxel is labelled with either blue or red colour if it is reproducibly assigned to the same eye on the two different occasions. Thus, the maps shown depict the patterns induced by stimulation of one eye versus the other, as well as their reproducibility in a single subject on two different occasions. Adapted from Yacoub et al. [58].

Figure 9.

Mapping frequency selectivity in the auditory pathway. Adapted from Formisano et al. [78], De Martino et al. [68], Moerel et al. [67], and De Martino et al. [40].

Figure 10.

(a) Standardized population-tuning curve in the human superior parietal lobule (SPL) in a maze-path direction task shown together with the standardized population-tuning curve obtained with single cell electrophysiology recordings in the posterior parietal cortex of monkeys performing practically identical tasks. All tuning curves were standardized with respect to their range, aligned to their maximum and averaged across voxels or cells. (b) The correlation between the actual maze-path direction and that predicted by the population vector calculated from MR voxels. Adapted from Gourtzelidis et al. [87].

Figure 11.

Comparison between activation patterns observed with task-fMRI when subjects are performing a simple ‘hand task’ with the right- or left-hand and ICA components extracted from the 3 T resting-state fMRI data from the Human Connectome Project (HCP) database. Patterns mapped onto the group-average inflated cerebral and the cerebellar atlas surface that has been mapped to the MNI atlas stereotaxic space. Adapted from Van Essen et al. [104].

Figure 12.

Schematic diagram of the approximate spatial and temporal scales of measurements for multi-photon imaging with calcium and voltage reporters and fMRI, showing that, with impending technological advances, one can aim to overlap these two methods so that it will be possible to perform recordings of activity of each individual neuron in an computational ensemble (e.g. in a cortical column) in an animal model while fMRI can detect the activity at the level of the same ensemble but over the entire animal or human brain (i.e. cover the entire brain at the sub-columnar spatial resolution sufficient to obtain images of the columnar activity).