Abstracts

The optical properties of brain tissue change when neurons fire action potentials. These activity-evoked optical changes involve alterations in the light scattering and absorption properties of the tissue, and are known as [lsquo]intrinsic optical signals[rsquo] (IOS). IOS is thought to be generated by a combination of at least three distinct physiological mechanisms: i) changes in blood volume, ii) changes in blood oxygenation, and iii) blood-independent changes associated with ion-fluxes and volume changes of neurons and glia. The usefulness of IOS as a clinical and research tool depends upon an understanding of the relationships between neuronal activity and cerebral hemodynamics, and how these relationships are represented in IOS data. The goal of this study was to better understand these relationships.
The two major mechanisms that mediate activity-evoked hemodynamic changes are i) increases in blood volume mediated by dilation of the smallest pial arterioles lying on the cortical surface, and ii) increases in blood oxygenation in the veins draining regions of activated cortex. Our experimental strategy was to acquire IOS data at various optical wavelengths, and at sufficiently high-magnification to resolve the changes occurring within these distinct microvascular compartments. Images were acquired using a CCD camera with a high dynamic range (12 bits), at 100 ms/image and a spatial resolution of 1 um/pixel at the highest magnification.
One set of studies focused on determining optical wavelengths that were selective for changes in blood oxygenation or blood volume. By surveying the visible spectrum, we found that light at 535 nm (+/- 5 nm ) was highly selective for changes in blood volume, whereas light at 660 nm (+/- 5 nm) was selective for changes blood oxygenation. We found that at 535 nm, activity-evoked optical changes were restricted mostly to the smallest pial arterioles, and were precisely correlated to changes in the diameter (and hence volume) of the vessel. At 660 nm, optical changes were restricted mostly to the veins draining activated cortex. A second set of studies focused on determining the relationships between neuronal activity and changes in blood volume and blood oxygenation. By combining optical imaging and electrophysiological recordings, we found that blood volume maps correlated accurately to the spatial locations of firing neurons. However blood oxygenation changes were at their maximum in the larger draining veins. These observations were consistent in both functional and epileptiform (interictal and ictal) activity.
These results demonstrate that i) IOS imaging can selectively map changes in either blood volume or blood oxygenation during normal and epileptiform activity, ii) changes in pial arteriole diameter underlie the IOS changes at 535 nm, and increases in blood oxygenation in veins underlie the IOS changes at 660 nm, and iii) blood volume IOS maps provide the most accurate spatial localization of neuronal activity.