Synopsis

Keywords: Contrast mechanisms: fMRI, Neuro: Brain function, Neuro: Brain connectivity

Laminar fMRI has been increasingly used for determining feedforward and feedback inputs to the cortex. To properly design and interpret laminar-resolution fMRI experiments, it is critical to examine biophysical and physiological sources of hemodynamic and fMRI signals. In this education talk, we will discuss 1) laminar fMRI contrasts such as GE-BOLD, SE-BOLD, CBV and CBF, 2) up-to-dated depth-dependent neurophysiological findings, and 3) relationships between laminar-specific electrophysiology and hemodynamic responses.

High-resolution fMRI at ultrahigh fields is getting popular to map functional sites and connectivity with high sensitivity and specificity. To advance brain research further, it is critical to dissect different circuits, such as ascending feedforward and descending feedback circuits, with fMRI. Since cortical depth-dependent fMRI (often referred to as ‘layer fMRI’, ‘laminar fMRI’) may determine layer-specific neural circuits, it has been of great attention to fMRI scientists (see https://layerfmri.com/category/layer-fmri-papers/). However, fMRI is an indirect measure of neuronal responses via the vascular or hemodynamic responses, thus it is critical to understand the relationship between laminar fMRI and physiology for designing fMRI studies and interpreting the laminar fMRI signal properly (e.g., review articles of neurophysiological basis: Poplawsky et al., 2019; Self et al, 2019).

Initially, we need to review existing fMRI contrasts in terms of vascular specificity and sensitivity (see Fukuda et al., 2021). The vasculature is hierarchically connected so that local responses will have up- and downstream effects on vessels far from the site of neuronal activity. Because the utility of laminar-resolution fMRI is contingent on separating these neural-specific responses from the nonspecific, many different forms of fMRI contrasts have been examined (Jin & Kim, 2008a). The most popular approach is GE BOLD due to its high sensitivity, high temporal resolution and whole-brain coverage. However, this high sensitivity is partly due to the contributions of non-specific draining vessels, including pial veins. Thus, it is crucial to remove non-specific contributions to improve spatial specificity with differential analysis (Olman et al., 2012) and model-basis deconvolution (Markuerkiaga et al., 2016; Marquardt et al., 2020). Alternatively, the SE-BOLD approach refocuses dephasing signals around the large vessels, improving spatial specificity to microvessels. Thus, SE-BOLD fMRI is a better approach for laminar fMRI (Zhao et al., 2004; Han et al., 2021), compared to GE BOLD, but with reduced sensitivity.

Since the majority of CBF originates from capillaries, fMRI with arterial spin labelling was shown to have improved neural specificity compared to GE BOLD and to overlap well with layers that have increased synaptic activity (Duong et al., 2000). However, CBF fMRI has a lower sensitivity compared to BOLD fMRI. Similar to CBF, the CBV contrast also provides better laminar specificity compared to GE-BOLD fMRI (Zhao et al., 2006), since it originates from arterial macrovessels and microvessels. Consequently, noninvasive CBV-weighted vascular space occupancy (VASO) fMRI-based techniques (Huber et al., 2017; Jin & Kim, 2008b; Lu et al., 2003) are increasingly adopted for laminar fMRI.

Next important point is the relationship between laminar neural activity and hemodynamic responses. The most intuitive model for laminar neurovascular coupling studies is the primary sensory cortex such as V1 and S1. Sensory cortices generally consist of six layers; however, these layers typically function as three. Sensory stimulation evokes a transient change in neuronal activity first in layer IV (i.e., granular layer) and, subsequently, activity spreads immediately into the supragranular (layers I – III) and infragranular (V and VI) layers, and recruits recurrent and feedback circuitry to produce sustained neuronal activity across all the layers. Indeed, sensory stimuli induced the highest fMRI response in layer IV where has the highest local field potentials and multiunit activity (Self et al., 2019; Jung et al., 2022). Similar correspondence between CBV and electrophysiology was observed in the olfactory bulb (Poplawsky et al., 2015). These suggest that fMRI can identify the synaptic input layer.

Many important unresolved issues (discussed in Fukuda et al. 2020) are 1) why the first synaptically activated layer induces the highest CBV and CBF response, even though there are recurrent circuits among layers, 2) why the late CBV response is more specific to the synaptic active layers (Jin & Kim, 2008a), 3) why the hemodynamic response time difference is related to laminar processing (Jung et al., 2021; Petridou & Siero, 2019; Silva & Koretsky, 2002), and 4) whether neural output layers can be identified from synaptic input layers (Huber et al. 2017). These issues will be discussed.

Acknowledgements

Supported by Institute for Basic Science in Korea (IBS-R015-D1)