UCL Logo

Hemodynamic Signals and Local Neuronal Processing

Ralf Galuske

Max Planck Institute for Brain Research, Germany

Functional brain imaging methods such as functional MRI (fMRI) and optical imaging of intrinsic signals detect neural activity indirectly by measuring activity dependent variations in the level of blood oxygenation. The exact neural basis of this blood-oxygenation level dependent (BOLD) signal is still controversial. However, there is increasing evidence that BOLD-signals are more closely related to local field potentials (LFPs) than to action potentials. LFPs reflect mainly synaptic activity and their amplitude strongly depends on the temporal coherence of synaptic events. This coherence is particularly high when neuronal networks engage in synchronous oscillatory activity. In the present study we examine the relationship between BOLD and neural activity by simultaneously recording electrical and hemodynamic responses in cat visual cortex. Increasing stimulus intensity enhanced spiking activity, high frequency LFP-oscillations and the BOLD-response whereby the BOLD-signal correlated best with the power of LFP-oscillations in the high gamma frequency range (50-90 Hz). With constant stimulus intensity we observed spontaneous variations of BOLD-responses which showed no clear relation to spike rate variations but were tightly correlated with the power of gamma frequency LFP-oscillations. These findings imply that the amplitude of the hemodynamic response reflects more closely the temporal coherence of neuronal population activity than the rate of action potentials suggesting novel interpretations of the relation between BOLD-signals and neuronal processing.