Negative BOLD Research Articles

Neuronal basis of neuroimaging in motor networks

In attempting to understand the physiology of the brain, it is most helpful to combine a number of imaging and physiological techniques. A multimodality approach with fMRI, EEG and transcranial magnetic stimulation (TMS) allows a comprehensive non-invasive approach. There needs to be particular caution using the BOLD technique for fMRI since it is so far removed from neuronal events. Recent demonstrations are reassuring that show perfusion technique results to be similar to BOLD results.1 The main executor of movement in the cortex is the contralateral primary motor cortex (M1) and generally it shows the largest BOLD signal in fMRI experiments. The timing of M1 activity can be seen with the latter part of the movement-related cortical potential (MRCP) with EEG measurements. M1 activity is greater with more frequent movements, movements with greater amplitude and with greater force. In no-go trials of go/no-go experiments, there is no fMRI activity in M1 suggesting that nothing is happening.2 However, there is similar EEG activity compared with go trials, and TMS studies show active inhibition,3} suggesting that there is inhibitory activity in M1, but that this does not reveal itself with fMRI. The failure of fMRI to show change in this circumstance has been interpreted as due to the fact that inhibition does not take as much metabolic energy as excitation. On the other hand, prolonged suppression of a motor action does show a negative BOLD response (as well as TMS inhibition), suggesting that with inhibition over a longer time, the activity in the network is diminished.4 Some fMRI studies show an increased activity in the ipsilateral motor cortex while others show a decrease. EEG shows similar activity contralaterally and ipsilaterally in the early part of the MRCP and less activity ipsilaterally in the latter part of the MRCP. TMS can show excitation or inhibition depending on the motor task. Detailed analysis of the fMRI signal shows that increases likely derive from premotor cortex and decreases from M1.5 Hence, ipsilateral M1 may be inhibited and ipsilateral premotor cortex excited.

SU-FF-I-32: Detectability of Vasodilatation During Breath Holding by Dynamic CBV-Based MRI: Comparison with BOLD-Based MRI

Purpose: To investigate the sensitivity of a CBV-based MR technique, vascular space occupancy (VASO) imaging, in the detection of vasodilatation during different durations of breath holding (BH) and its clinical feasibility as compared with BOLD MRI. Method and Materials: Experiments were performed on a 1.5T Siemens MRI scanner. For VASO, a non-slice-selective IR-GE-EPI was used with TR/TE/TI = 2000/9.3/665 ms, SW = 6 mm and in-plane resolution = 3 mm. For BOLD, a GE-EPI was used with TR/TE = 2000/ 60 ms. Two normal volunteers each performed 4 experimental runs, including two-cycle 30-s and 15-s BH each for BOLD and VASO. CNRs of were determined from ROIs drawn within the thalamus. A similar protocol was also carried out in a patient with right frontal anaplastic astrocytoma to review clinical feasibility. Results: Positive BOLD signal changes were noted during 30-s and 15-s BH (4.9 and 2.2 %), and the CNRs were 4.0 and 2.5. VASO method was able to detect CBV related negative signal changes (−3.1 and −1.8%, for 30-s and 15-s BH), and the CNRs were 2.3 and 1.8. Initial analysis of the patient data set revealed negative intra-tumoral BOLD signal changes. While the global pattern of the VASO map agreed well with BOLD, the extent of intra-tumoral changes was smaller. The BOLD map was then re-thresholded to a lower p-value which demonstrated intra-tumoral changes similar to those with VASO. Conclusion: CNRs of VASO were lower than those from BOLD. However, they were closer to BOLD (58 and 72% for 30-s and 15-s BH) than previously published results from visual stimulation (∼ 30%). Discrepancies between two techniques in the clinical case were likely resulted from the differences in sensitivity. However, given that CBV-based signal change is more directly related with the vasomotor response, it could potentially be more clinically relevant.

Long-range EEG phase synchronization during an arithmetic task indexes a coherent cortical network simultaneously measured by fMRI

An open question lies in whether or not distributed activities in the distant brain regions are integrated into a coherent ensemble for cognitive information processing. Long-range phase synchronization is often observed by scalp EEG measurements during cognitive tasks and is considered to provide a possible neural principle for the functional integration of distributed neural activities. Synchronization could be reflected at the neuron firing level or at the local field potential and could appear in the scalp EEG under certain conditions on neural spatial and temporal coherence. To examine if phase synchronization is concerned with the integration of distant regions, we proposed a method to extract brain activities associated with task-dependent phase synchronization by combining simultaneous fMRI and EEG. By applying this method in a mental arithmetic task, we found a dominant task-dependent increase of phase synchronization around 14 Hz (in beta frequency) across bilateral parietal sites that were associated with both negative and positive BOLD responses. Functional connectivity analyses of these regions demonstrated that an increase in hemispheric beta synchronization was associated with a linking between the cross-hemispheric regions (left angular gyrus and right superior parietal gyrus) and also among the anterior–posterior regions (right dorsolateral prefrontal cortex, putamen, and right superior temporal gyrus). These findings indicate that the positive BOLD regions (dorsolateral prefrontal cortex and superior parietal lobule) are linked with other negative BOLD regions. We also discussed the possible importance of beta synchronization in the formation of a working memory network.

Decreases in neuronal activity and negative BOLD response in non-stimulated regions of monkey V1

Decreases in neuronal activity and negative BOLD response in non-stimulated regions of monkey V1

A long-range cortical network emerging with theta oscillation in a mental task.

Human scalp EEG has demonstrated that global coherence among distant areas increases during cognitive tasks, suggesting that oscillating neural activities work to generate global neuronal assemblies for cognitive functions. The theta oscillation in a frequency range of 4-8 Hz with large amplitudes which emerges during mental tasks around the frontal midline region is called fm theta. If theta oscillation concerns the global neuronal assemblies, fm theta should be associated with regional activities that depend on task conditions. In the present study, we examine the EEG-related brain activities by developing simultaneous EEG and fMRI during a mental calculation task. EEG-related negative BOLD was dominant over anterior medial regions, suggesting a major contribution of negative BOLD to fm theta. Negative and positive BOLD were found over distant regions. Functional connectivity analyses revealed that the connectivity varied remarkably according to mental conditions. In the rest condition, the connectivity was localized, whereas in the task condition, a long-range coherent network was formed by the anterior midline, posterior cingulate and right middle temporal cortices with linking between the right middle temporal and left lateral cortices during numerical processing. Further EEG analyses indicate that the long-range coherent network executing cognitive functions is coordinated in the time window of theta oscillations.

BOLD signal sign and transient vessels volume variation.

The purpose of this work was to investigate the relation between BOLD signal sign and transient vessels volume variation induced by apnea. This stimulus consisting of breath holding after inspiration is able to induce a light slowing down in venous blood flow like in a sort of Valsalva maneuver. We observed diffuse negative BOLD responding areas at cortical level and a stronger negative response in correspondence of the main sinuses. These phenomena seem to be unrelated to a specific neural activity, appearing to be expressions of a mechanical variation in the hemodynamics. Our study suggests that particular care must be considered in the interpretation of fMRI findings, especially when patients with vascular-related cerebral diseases are involved.

Sustained Negative BOLD, Blood Flow and Oxygen Consumption Response and Its Coupling to the Positive Response in the Human Brain

Most fMRI studies are based on the detection of a positive BOLD response (PBR). Here, we demonstrate and characterize a robust sustained negative BOLD response (NBR) in the human occipital cortex, triggered by stimulating part of the visual field. The NBR was spatially adjacent to but segregated from the PBR. It depended on the stimulus and thus on the pattern of neuronal activity. The time courses of the NBR and PBR were similar, and their amplitudes covaried both with increasing stimulus duration and increasing stimulus contrast. The NBR was associated with reductions in blood flow and with decreases in oxygen consumption. Our findings support the contribution to the NBR of (1) a significant component of reduction in neuronal activity and (2) possibly a component of hemodynamic changes independent of the local changes in neuronal activity.

Effects of intracellular pH, blood, and tissue oxygen tension on T1ρ relaxation in rat brain

The effects of intracellular pH (pH(i)), paramagnetic macroscopic, and microscopic susceptibility on T(1) in the rotating frame (T(1rho)) were studied in rat brain. Intracellular acidosis was induced by hypercapnia and pH(i), T(1rho), T(2), diffusion, and cerebral blood volume (CBV) were quantified. Taking into account the CBV contribution, a prolongation of parenchymal T(1rho) by 4.5% was ascribed to a change in tissue water relaxation caused by a one unit drop in pH(i). Blood T(1rho) was found to prolong linearly with blood oxygenation saturation (Y). The macroscopic susceptibility contribution to parenchymal T(1rho) was assessed both through BOLD and an iron oxide contrast agent, AMI-227. The T(1rho) data from these experiments could be described by intravascular effects with insignificant effects of susceptibility gradients on tissue water. Tissue oxygen tension (PtO(2)) was manipulated and monitored with microelectrodes to assess its plausible contribution to microscopic susceptibility and relaxation. Parenchymal T(1rho) was virtually unaffected by variations in the PtO(2), but T(1) was shortened in hyperoxia and T(2) showed a negative BOLD effect in hypoxia. It is demonstrated that pH(i) directly modulates tissue T(1rho), possibly through its effect on proton exchange; however, neither BOLD nor PtO(2) directly influence tissue T(1rho). The observations are discussed in the light of physicochemical mechanisms contributing to the ischemic T(1rho) changes.

Negative BOLD signal suggests functional recruitment of blood flow

Negative BOLD signal suggests functional recruitment of blood flow

Negative BOLD response and its coupling to the positive response in the human brain

Negative BOLD response and its coupling to the positive response in the human brain

Effect of pentobarbital on visual processing in man.

To investigate the effect of sedative agents on visual processing in humans, we analysed the BOLD contrast signal response to a visual stimulation paradigm in 15 healthy, adult volunteers using functional magnetic resonance imaging. The subjects were tested during alert state and under sedation following intravenous administration of pentobarbital. The injection of pentobarbital not only significantly reduced the response signal strength but the reduction in BOLD contrast signal was related to the ratio of amount of sedative administered and the subject's body weight. The three subjects with the highest relative sedative dosage even displayed an inverted (negative) BOLD contrast signal. A significant reduction in the number of positively correlating pixels was found 15 min after administration of pentobarbital. All measured parameters returned to near pre-sedative levels by the end of the experimental session. The relative dosage dependence of the strength of the BOLD signal the negative BOLD signal in the three subjects with the highest relative sedative dosage indicates that pentobarbital had a more pronounced effect on cerebral blood flow than on cerebral oxidative metabolism.