Table 2.
Task based fMRI-neural activity comparisons with fMRI and neural activity measures obtained in different subjects with the same task(s).
| Article | Cortex sampleda | Neural activity measureb | Taskc | Comparison | Findings | Comment |
|---|---|---|---|---|---|---|
| Huettel et al., 2004; Engell et al., 2012 | TO | ECoG visual EP (9) | Visual checkerboard of varying durations | Changes in magnitude, form of EP vs. BOLD with varying durations at different sites | Inconsistent BOLD-EP relation: BOLD monotonic, non-linear increase with duration in O, inferiorT; EP: O sustained, inferior T phasic. No duration or power changes | Attempt at direct comparison to findings of Logothetis et al. (2001) in monkey. Concluded that EP-BOLD relation differed across brain regions. See text for recent reanalysis of non-phase-locked ECoG power |
| Mukamel et al., 2005 | Auditory | ME depth (2) | Viewing movie | Correlation unit firing, LFP power 1–130 Hz convoluted with standard hemodynamic response function, and BOLD | Linear correlation positive with unit firing, (r = 0.39, 0.5) LFP 40–130 Hz, (r = 0.39, 0.46) negative 5–15 Hz. Unit activations colocalize with fMRI | Often cited as indicating unit firing rate related to BOLD signal magnitude. Unit and LFP 40–130 Hz activity confounded |
| Meltzer et al., 2008 | FTPO including midline bilateral | ECoG (14) | Sternberg working memory, comparing varying load | Visual colocalization of fMRI, LFP power significant changes, pooled by lobe | 30–50 Hz up with increased load all O sites. 5–13: up med F, down O, lat F. BOLD little over-lap except increase colocalized L precentral only | |
| Nir et al., 2007 | Auditory bilateral | ME depth (3) | Viewing movie | Convoluted unit firing, LFP 40–130 Hz to hemodynamic response function. Correlations during viewing | Unit firing correlated with BOLD (r = 0.70) only when correlations between unit firing of nearby neurons present (r = 0.66). 40–130 Hz correlated with BOLD r = 0.62 | BOLD correlation with unit firing only when unit firing inter-correlated |
| Privman et al., 2007 | O, posterior lateral T | ECoG (7) | Viewing movie, object categorization faces, houses, tools | Convoluted ECoG 8–13, 13–30, 30–70 HZ to hemodynamic response function. Correlations during viewing. | 60% of electrodes with significant 30–70 Hz correlated (r = 0.34) with BOLD in O, lateral T auditory cortex category selectivity of best ECoG electrodes > that of best BOLD | No data on selective category co-localization of ECoG and BOLD |
| McDonald et al., 2010 | L inferior F, ventral TO, ST. Regions selected those with BOLD change | ECoG (6) | Primed vs. unprimed words | Randomization test of colocalization of significant ECoG 70–190 Hz power change at sites with significant BOLD change | Significant in all subjects at ventral OT, in only part for other areas. Concluded “tightly coupled” | No data on sites without fMRI change. Also compared to scalp magneto-encephalo-gram |
Notes for row 1:
a
F, frontal; T, temporal; P, parietal; O, occipital; L, left; R, right.
b
ECoG, electrocorticogram, unless indicated from subdural grids. EP, evoked potentials from ECoG. ME, microelectrodes, with local field potentials (LFP) and single neuron (units) action potentials. “depth” indicates recordings from depth electrodes. Number of subjects indicated in parentheses.
c
vs., verses. Except as noted fMRI and neural activity task essentially the same.