Abstracts

Rationale: Previous research has suggested that optical changes, specifically changes in near-infrared optical scattering, may precede electrographic seizure onset in in vivo models. However, previous techniques have required craniectomy for direct cortical imaging or the use of invasive implantable fiberoptic probes. In this work, we describe the first adaptation of in vivo optical coherence tomography (OCT) imaging through a thinned-skull preparation to optically assess rodent cortical seizures. Methods: 6-8 week old mice were anesthetized with ketamine/xylazine and a 4x4 mm thinned-skull cortical window was created with a dental bur over the cerebral cortex. The cerebral cortex was then imaged continuously with our OCT system. The spectral domain OCT system we developed contains a broadband, low-coherence light source consisting of two superluminescent diodes (SLD) centered at 1295 nm with a bandwidth of 97 nm resulting in an axial and lateral resolution of 8 μm and 20 μm respectively. The brain was continuously scanned creating a 3 mm imaging plane with an imaging depth of 2 mm. Axial line scans (A-lines) were acquired at 15 kHz and each cross-sectional area consisted of 2048 A-lines. A 10 min. imaging baseline was collected before a saline injection was administered. After 35 to 40 min, pentylenetetrazol (PTZ) (100mg/kg, i.p.) was injected to induce cortical seizure activity. Mice were observed for whisker twitches/myoclonic jerks and until a stage 5 seizure occurred with fore- and hindlimb clonus. Results: To analyze the changes in backscatter intensity, we calculated average intensities from cortical regions of interest (ROIs) over time. Two controls were used: the first was a 10 min baseline before any injection was administered to assess optical stability and the second was an injection of saline with a volume equivalent to the PTZ dosage. We observed no significant change in intensity (95% confidence interval) from baseline after saline injection. However, in all experiments (n=4), decrease in backscattered intensity after PTZ injection was observed. Latency to two standard-deviation decrease in optical scattering (empirically defined as optical "threshold") occurred <10 min after PTZ injection, whereas latency to whisker twitches/myoclonic jerks was >10 min and latency to generalized seizure activity was >20 min. Conclusions: We have adapted the technique of OCT for in vivo imaging of the rodent cortex before and during PTZ-induced seizure activity. Stable baseline images through the thinned-skull preparation were uniformly obtained, and no significant changes were seen following saline injection. Reduction in backscatter intensity was observed following PTZ injection but prior to generalized seizure activity. These results provide proof-of-principle for the use of OCT through a thinned-skull preparation to monitor cortical optical changes with high spatial and temporal resolution in vivo.