|Home | About | Journals | Submit | Contact Us | Français|
Epilepsy results from “electrical storms” inside the brain that cause recurring seizures. About 2 in 100 people in the United States experience an unprovoked seizure at least once in their lives. The population and shape of involved epileptic neurons vary along the time course of each epileptiform event in a very short time span. Current clinical functional imaging methods, such as functional magnetic resonance imaging (fMRI), positron emission tomography (PET) and single-photon emission computed tomography (SPECT), are limited by their low temporal resolution in documenting such paroxysmal epileptiform events. High temporal resolution is critical to overcome the motion artifacts caused by patients or procedures during epilepsy surgery. Although optical mapping methods such as intrinsic optical imaging (IOS) (Bahar et al,. 2006; Inyushin et. al., 2001; Haglund and Hochman, 2004; Schwartz et al., 2004; Schwartz and Bonhoeffer 2001) and voltage-sensitive dye imaging (VSD) (Cohen et. al,. 1986) have limitations in imaging depth and tissue toxicity, the development of epileptic animal models has made them very useful (Raol et. al., 2012). More recent developments in the optical mapping of neural activities include the optical coherence tomography (OCT) (Satomura Y. et. al., 2004; Aguirre A. D. et. al., 2006; Rajagopalan U. M. et. al., 2007; Chen Y. et. al., 2009; Sato M. et. al., 2010; Liang et. al., 2011; Lenkov et. al., 2012; M. Eberle et. al., 2012) and photoacoustic microscopy (PAM) (Hu et. al., 2009; Maslov et. al., 2008; Tsytsarev et. al., 2011; Wang et. al., 2003; Wang 2008; Wang 2009; Liao et. al., 2010; Liao et. al., 2012a; Liao et. al., 2012b). OCT can image both the refractive index modulation of the tissue that surrounds epileptic neurons with optical scattering contrast and the neuron-vasculature coupled blood flow patterns with Doppler OCT contrast. PAM represents an innovation in the functional imaging of red blood cells and blood flow. It provides functional data such as the oxygen saturation of red blood cells in the microvasculature. Because PAM is based on pure optical absorption contrast, it does not require the dense optical scans required by Doppler OCT, and it is immune to axial motion artifacts that may significantly compromise the blood flow images generated by the Doppler OCT in in vivo applications. As a rapidly evolving imaging technology, PAM promises deep imaging into tissue due to its multi-scale imaging capability (Wang et. al., 2012). By recording the optical refractive index modulation with OCT technology and simultaneously documenting the hemodynamic changes in epileptiform events with PAM, we can, for the first time, perform epilepsy mapping with high temporal and spatial resolution and dual optical contrasts. We expect our demonstrated technology to have great utility in applications such as epilepsy drug development and epilepsy surgery.
For each experiment, a Swiss Webster mouse (Hsd: ND4, 25–30 g; Harlan, Indianapolis, IN) was anesthetized by an intraperitoneal (IP) injection of a mixture of ketamine (87 mg/kg) and xylazine (13 mg/kg). The anaesthetized animal was placed in a custom-made stereotaxic head holder, and the left dorsal portion of the skull was exposed by surgically removing the scalp and muscle. A cranial opening (~4–5 mm2) was made using a dental drill over the left hemisphere, and the exposed dura mater surface was cleaned with artificial cerebrospinal fluid (ACSF). Throughout the experiment, the animal was supplied with breathing-grade compressed air (AI B300, Airgas, MO) and maintained under anesthesia using isoflurane (1.0–1.5% with an airflow rate of ~1 L/min), while the body temperature of the animal was maintained at 37 °C by a temperature-controlled electrical heating pad. After each experiment, the animal was euthanized with an overdose of pentobarbital. All experimental animal procedures were carried out in conformance with the laboratory animal protocol approved by the Animal Studies Committee of Washington University in St. Louis.
After craniotomy, 0.35 μL of a 25 mM solution of 4-aminopyridine (4-AP) in artificial cerebrospinal fluid (ACSF) was injected into cortical layers II–III, using an injector device (Nanojet II) with a 15–25 μm diameter glass microcapillary (Bahar et al. 2006; Tsytsarev et al. 2011). The injector was mounted on a micromanipulator that allowed injections 0.2 – 0.3 mm below the dura mater surface. A single channel electroencephalogram (EEG) was recorded through a screw-type electrode placed in the left hemisphere in the skull, then amplified by an AC/DC differential amplifier (A-M Systems, Model 3000), digitized at 5000 Hz, and recorded. The start of EEG data acquisition was manually initiated.
A dual modality optical imaging system that can simultaneously provide a co-registered OCT image and a PAM image (Figure 1) was used to provide optical mappings of epilepsy. The OCT imaging channel provided depth-resolved brain tissue structure images based on optical scattering contrast (Rao et. al., 2008). The PAM channel provided depth-resolved brain vasculature images based on optical absorption contrast (Rao et. al., 2010a). OCT and PAM Alines were acquired sequentially for each scanning location. Thus, an OCT image and a PAM image were recorded virtually simultaneously. Each B-scan frame contained 800 scanning points (A-lines). At an imaging speed of 5,000 A-lines-per-second, 500 B-scan frames were acquired within an 80-second time window (Rao et. al., 2010b). Similar to the procedures described in our previous publications (Tsytsarev et. al., 2011; Hu et. al., 2009), we identified two vessels close to the 4-AP injection site and performed repeated B-scans over the selected vessels to map epilepsy in both spatial and time dimensions. The temporal resolution of epilepsy mapping was 6.25 Hz. The spatial lateral resolutions of epilepsy mapping with OCT and PAM were 5.2 μm and 3.5 μm, respectively.
After image acquisition, three types of image data (EEG, OCT, PAM) were analyzed separately. The EEG data was simply displayed as voltage versus time. For the OCT data, the optical scattering signal intensity was calculated by Fourier transformation of the interference fringe recorded by a line-scan CCD camera. Using the first B-scan image as a baseline, we calculated the depth-resolved reflectivity change ΔR(x,t), where x represents depth and t represents the time or B-scan number, by averaging 800 A-lines within every B-scan image. Then, ΔR(x,t) was averaged along the imaging depth to generate an averaged reflectivity change over time, ΔR(t). For PAM data, the A-line amplitude of the photoacoustic signal was extracted via Hilbert transformation. The vessel diameter was directly shown by the maximum amplitude projection (MAP) of PAM B-scan images. Changes in vessel diameter could be directly visualized from the plot of MAP images along the time axis.
After a 4-AP intracortical injection, epileptic seizures occurred periodically for 2 – 4 h, lasting 20 – 200 s at intervals of 2 –20 min. Figure 2A records a typical seizure EEG signal course lasting about 80 seconds. Figures 2B and 2C plot the 1D vertical MAP images from PAM along the horizontal time axis. A large vasodilatation of the blood vessels (veins), as shown in figure 2B, signals the electrographic onset of seizure, which is well-correlated with the EEG signal. The vessel (artery) in figure 2C displays a similar vasodilatation effect, but with reduced amplitude and prolonged duration. It is interesting to note that significant vasodilatation occurred only near the injection site. Outside a cortical area of about 1 mm2, vasodilatation was below our detection threshold of about 10%. Additionally, it seems that the vasodilatation may signal the beginning of seizure earlier than conventional EEG signals. From figure 2B, we observed that after the large vessel dilation the vessel was smaller than the baseline diameter. Further study might be interesting, but is beyond the scope of this communication.
The depth-resolved reflectivity change ΔR(x,t) during seizure is shown in figure 2D. From the plot of ΔR(x,t), we can easily visualize the depth-resolved reflectivity change along the whole time course of the seizure. A reflectivity change ΔR(t) averaged along the depth direction at each sampling time point can represent the trend of reflectivity variation in a simple plot, as shown in Figure 2E. The suddenly decreased reflectivity (represented by the color change from green to blue in Figure 2D and by the curved dip in Figure 2E) coincided with the start of vessel dilation. The changes in cortical reflectivity ended approximately at the seizure’s end. In spite of the low transparency of the brain tissue, we were able to record the tissue-scattering signal from 2 mm below the surface, which is deeper than the cortical thickness. With fair confidence, we conclude that both ΔR(x,t) and ΔR(t) are more sensitive in signaling the start and the end of the seizure than EEG signals.
The 4-AP model of epileptic seizures (Bahar et. al., 2006; Tsytsarev et. al., 2011; Zhao et. al., 2011; Raol and Brooks-Kayal, 2012) allows us to investigate periods of induction, maintenance, and propagation of seizure discharges, and has been extensively studied using different optical methods. It is generally known that epileptic seizures are accompanied by a local increase in cerebral blood flow to the epileptic focus (Zhao et. al., 2011, Hirase et. al., 2004; Santisakultarm et. al., 2011; Schwartz et. al., 2004), but the relationship between seizures and neuronal activity remains unclear. The trigger mechanism of the seizure as a synchronized activity of the neural network is also not very well established. Although it is unlikely that vasodilatation itself can initiate the seizure, fast vasodilatation might be a first indicator of the local biochemical process that accompanies the seizure’s beginning. It was proposed by de Vasconcelos et. al., (1995) that nitric oxide triggers vasodilation in response to focal epileptic seizures, but in their work, vasodilatation was observed in a relatively large area, including both the cortex around epileptic foci and also (bilaterally) the substantia nigra and the parafascicular thalamic nucleus. It seems possible that astrocytes’ release of glutamate in the area of the epileptic foci plays a causal role in synchronous firing of a large neural population (Carmignoto and Haydon, 2012; Inyushin et al, 2012). Synchronous activity of the astrocytic syncytium can cause changes in the optical features of the cortical tissue that can be observed by OCT.
OCT has been used for in vivo brain imaging in animal experiments (Satomura Y. et. al., 2004; Aguirre A. D. et. al., 2006; Rajagopalan U. M. et. al., 2007; Chen Y. et. al., 2009; Sato M. et. al., 2010; Liang et. al., 2011; Lenkov et. al., 2012; M. Eberle et. al., 2012) and provides high spatial resolution and a wide field of view. The modulation of the optical refractive index by the epileptic neuronal activity during a seizure coincides with the observation of surface reflectivity changes. As demonstrated in our experiments, OCT represents a very effective direct optical epilepsy mapping method, with a depth range of about 2 mm. Due to the well-known neurovasculature coupling mechanism, dynamic blood flow changes such as the vasodilatation observed in our experiments may be established as another effective optical epilepsy mapping method for documenting epilepsy seizure. Theoretically, most blood flow imaging methods, including Doppler OCT and its variants, could be used to record the dynamic blood flow change caused by an epileptic seizure. The recently invented PAM technology has major advantages due to both its high contrast (as high as 40 dB) and its multi-scale imaging capability, which provides a much deeper imaging depth not achievable by other optical imaging technologies. In this demonstration, we used a very simple epilepsy mapping protocol, which does not utilize the full speed of our imaging systems (20,000 A-lines-per-second). It is very straightforward to upgrade the dual contrast imaging system to a 100,000 or 200,000 A-lines-per-second imaging speed, enabling more sophisticated mapping protocols. The dual contrast optical epilepsy mapping method demonstrated in this short communication presents a new tool for epilepsy drug development with small animal models (Raol et. al., 2012). With appropriate modifications, it could be used as an intra-operative tool for determining the boundary of epileptic tissue in epilepsy surgery (Haglund et. al., 2004).
We demonstrate for the first time in vivo epilepsy mapping with high spatial and temporal resolution and dual optical contrasts in an animal model. Through the variations in a depth-resolved optical coherence tomography signal with optical scattering contrast, we observed variations of the depth resolved optical reflection signal from epilepsy tissue. Simultaneously, through neurovascular coupling mechanisms and optical absorption contrast, we used photoacoustic signals to document the hemodynamic changes of the microvasculature surrounding the epileptic neurons. The epilepsy mapping results were confirmed by a simultaneously recorded electroencephalogram signal during epileptic seizure. Our new epilepsy mapping tool, with high temporal and spatial resolution and dual optical contrasts, may find many applications, such as in drug development and epilepsy surgery.
This work was sponsored in part by National Institutes of Health grants R01 EB008085, R01 CA134539, U54 CA136398, R01 CA157277, R01 CA159959, DP1 EB016986 (NIH Director’s Pioneer Award).. L.W. has a financial interest in Microphotoacoustics, Inc. and Endra, Inc., which, however, did not support this work. K.M. has a financial interest in Microphotoacoustics, Inc., which, however, did not support this work.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.