Although activity-related intrinsic changes in tissue reflectance associated with electrical activity or metabolism were first observed over 50 years, it was not until the 1980s that these intrinsic optical changes were used to map cortical activity in vivo. The increasing popularity of optical imaging is largely because this technique offers both high spatial and high temporal resolution simultaneously. The spatial resolution of intrinsic imaging is unparalleled among in vivo imaging techniques making it ideal for studying the fine functional organization of sensory cortices as well as the physiology of neurovascular coupling at the level of the arteriole, venule, and even capillaries. Because of these advantages, the number of studies using optical imaging has been growing rapidly.

• functional architecture of sensory cortices,
• investigating patterns of functional perfusion and neurovascular physiology,
• imaging development and plasticity, and
• characterizing the dynamic profile of disease processes.

Macknik and Haglund employed optical imaging to characterize the primate visual cortex. In their study they sought to characterize the functional architecture of the cortex. They were interested in finding out whether physical visual scene or the perception of the visual scene was the stimulus for visual system activation. In this study using optical imaging techniques, they determined that the visual system does not respond to the entire physical visual stimulus, but rather the visual perception. This type of study demonstrates the versatility and the diverse array of questions that can be answered using optical imaging of intrinsic signals.

• Characterizing functional perfusion and neurovascular physiology
  Intrinsic optical signals have multiple sources, including perfusion- and metabolism-related etiologies, making it a suitable modality for characterizing functional perfusion and neurovascular physiology. Optical imaging studies have been critical for describing the time course of functional perfusion-related changes, including changes in blood volume and hemoglobin concentration changes. In addition to characterizing the time course of functional signals, optical imaging has been used to determine the parameters which influence the magnitude of functional signals, including the absolute number of stimuli as well as the frequency of stimulation.
• Chronic imaging to characterize both developmental and plasticity-related changes
  In recent years several groups of researchers have used optical imaging forchronic imaging to characterize both developmental and plasticity-related changes in functional architecture and perfusion-related signals. The development of orientation domains has been characterized in ferret visual cortex, from the time that they are first observed to the time at which they are fully developed. The researchers showed that during this time (which is the peak of the critical period) the structure of the maps is remarkably stable.
• Imaging disease
  Optical imaging of intrinsic signals has recently been applied to imaging disease. One group of researchers reported imaging epileptiform afterdischarges following direct cortical stimulation of the human neocortex. They reported that activation patterns were dependent on the intensity and duration of the afterdischarge activity often found a pattern of alternating increased and decreased reflection, that they hypothesized to represent a central area of activation surrounded by areas of cortical inhibition or decreased activity.

• Neurovascular coupling and etiology of perfusion-based mapping signals
• Cortical spreading depression (CSD)
• Cortical working memory
• Cortical development and plasticity
• Intraoperative imaging and surgical guidance

OIS is ideally suited to investigate neurovascular coupling, as it allows imaging of perfusion-related responses with simultaneous monitoring of electrophysiological activity. We combine OIS imaging and spectroscopy with extracellular field potential recordings in rodent primary somatosensory cortex to investigate this interaction. Our results indicate that early deoxygenation changes in active cortex, related to the “initial dip” observed by some fMRI studies, correlate weakly with changes in the magnitude of underlying neuronal activity. Changes in cerebral blood volume (CBV), on the other hand, correlate robustly with changes in evoked neuronal activity. This difference may be partially attributable to the increased contrast-to-noise ratio (CNR) of the CBV response (~10x) relative to that of the early deoxygenation. Increased CNR also allows reliable mapping with less trial averaging, thereby reducing imaging time and motion-related noise contamination, especially in awake subjects.

Ideal functional mapping signals should not only correlate in magnitude with the degree of neuronal activity, they should also localize to activated cortical columns with high spatial specificity. The stereotypical organization of the rat whisker barrel cortex provides a convenient model system to study this question. We found that the CBV response localizes well to individual barrels, as visualized by cytochrome oxidase staining. Appropriate thresholding techniques, including a novel ANOVA-based approach we developed, provide high quality CBV-based maps with sub-millimeter columnar resolution.

Current projects involve combining multi-modal imaging with simultaneous electrophysiological measurements to assess changes in functional cerebral blood flow, volume, and oxygenation following CSD. The imaging techniques we utilize include multi-wavelength optical imaging, laser Doppler flowmetry, 2D laser Doppler imaging, and optical spectroscopy. Electrophysiological changes are monitored by recording evoked potentials with depth microelectrodes. Recent data suggest global blood flow and volume are depressed for1-2 hours following CSD induction. During this period, functional perfusion signals also show a significant attenuation reaching a relative minima at approximately 20 minutes post-induction followed by a gradual return to baseline. Evoked potentials show a much smaller attenuation and significantly faster recovery suggesting CSD produces a short-term (1-2 hrs) mismatch between changes in neuronal activity and functional vascular signals. In the future, we plan to examine the effect of age on recovery of neurovascular coupling following CSD. We also plan to extend out current paradigm to investigate neurovascular coupling stability following seizure.

Recently, in data collected in parietal cortex we observed that NIRS signals show a strong correlation to LFP responses in several different tasks of working memory. Interestingly, subtle differences in task performance apparent in the LFP data appear to be reflected in the NIRS signal. Future plans include making simultaneous measurements (NIRS and LFP) from multiple cortical sites to gain a better understanding of the interaction between different cortical areas during working memory.

Although many studies have investigated experience-dependent electrophysiologic, metabolic, and anatomic changes in the rodent barrel cortex, experience-dependent changes in functional perfusion have not been well characterized. Characterizing these changes is important because perfusion-related responses are used extensively to map brain function. Using a spared-whisker rodent model and OIS, we characterized the timecourse and spatial extent of experience-dependent changes in perfusion-related responses at different ages (2, 4, and 6 weeks). Imaging 1, 2 and 4 weeks after initial whisker plucking indicated that optical responses to spared whisker stimulation are subject to experience-dependent modulation, but that increases relative to age-matched controls do not occur until two weeks after initial whisker plucking in any age group. These changes are later than previous reports of electrophysiologic and metabolic changes in the spared barrel using the same model. We also found that experience-dependent changes in optical responses are reversible suggesting that perfusion-related responses are continuously modified by sensory experience. Furthermore, we report that the spatial extent of optical responses to stimulation of the spared whisker is similar to that for simultaneous stimulation of all whiskers in controls. These results have implications for studies of plasticity using perfusion-based functional neuroimaging techniques.

We conduct both pre- and intra-operative brain mapping experiments in order to answers questions about the basic neurovascular phsyiology of the human brain and in order to create better mapping tools for neurosurgical guidance. We use a multi-modality approach, comparing different brain mapping signals with the gold standard and with postoperative neuropschological testing also to better understand the functional organization of the human brain.

To investigate the similarities in etiology between BOLD fMRI signals and 610 nm optical imaging, we compared the spatial and temporal characteristics of the two responses in human patients during tongue activation. Several fMRI studies cite OIS findings to explain the etiology of fMRI signals, but few studies have been conducted using both modalities in the same subjects. Optical responses (measured intraoperatively) generally colocalized with BOLD fMRI responses (measured preoperatively), and were consistent with locations identified by intraoperative electrical stimulation mapping. We found that whereas fMRI activations were centered on and emphasized sulci (most likely because of the major contribution of large venous structures to the fMRI signal), OIS activation were more gyral. The timecourse of OIS activations were very similar to those observed for BOLD signals, appearing shortly after task onset and peaking approximately 6–8 s after task onset. A regression analysis of the timecourse data showed a significant linear correlation between the two signals.

We also determined the utility of preoperative fMRI to localize cortical areas subserving language function in neurosurgical patients. We found that the sensitivity and specificity of fMRI for mapping language areas depended upon both the task (expression or comprehension) and lobe (frontal, temporal, and parietal). Specifically, we found that the sensitivity and specificity of fMRI during expressive linguistic tasks was up to 100 and 66.7%, respectively, in the frontal lobe, and during comprehension linguistic tasks up to 96.2 and 69.8%, respectively, in the temporal and parietal lobes.