Non‐Invasive Cerebral Blood Volume Measurement during Epilepsy Using Multi‐Channel Near Infra‐Red Spectroscopic Topography

Abstract

Purpose: In scarch for noninvasive method for focus diagnosis we applicd near‐infixed spectroscopy (NIRS) which is a new technique for noninvasive nicasuremcnt of cerebral blood volume (CBV). We have previously confirmed the efficacy of this method in functional brain mapping using multi chaniiel NIRS during motor stimulation in humans. M(.thocls: Twclve paticnts with medically intractable epilepsy (I0 with tcmporal lobe epilepsy and 2 with parictal lobe epilcpsy) underwent the NIRS mapping. Seizure foci were determined by intracranial EEG i n all cases. NIR with 2 wavelengths of 7x0 and 840 nm was transmitted into the cranium from the scalp and its reflection was receivcd 30 min away from the transmitting probe. Probes making up X to 24 channels were mounted on the hcad shell (4 or 8 channels on each side) and CBV change was sequentially incastired in evcry I second before, during, and after the scizure. The epileptic seizure was induced wing betncgride injection, and the EEG monitoring and ictal SPECT wcre simultancously performed. Results: I n all cases, CVB increased significantly in the sidc of scizurc foci, which were dcfincd by intracranial EEG. The increase was ohscrved in 2 to 8 seconds after the onset of seizure and lasted for 20 to SO seconds. In 6 cases, the ictal SPCET rcvealcd hyperperfusion tireas in the foci.Discussition: Thc newinfrared lighl projected into the brain penetrates and rcflects randomly in the tissue. Thc ne;ir‐infrarcd light measured at some distance represcnts the absorption by qupathin the hrain tissue connecting the transmitting and the receiving probes. The depth of the measuring point depends on thc distance between the transmitting and recciving probes. It is reportcd that NIRS signal mainly rcflects thc absorption at the dcpth of I .2 to 2.0 cm from the scalp when the inter‐probe distance is 2.7cm. Because the human cerebral cortex usually lies about 10 to 20mm deep froin the scalp, the suitablc inter‐probe distance should be 25 to 30 mm to measure the activities of the cerebral surface. The inter‐probe distance was dctermined to he 30mm in the present study. Several researchers reported on the tight coupling between the neuronal activity and the rCBF increase. This was demonstrated and proved in animal experiments. In humans, PET and fMRI documented the rCBF increasc during neural activities in the course of the sensory stimulation of fingers. A strip of cortex was activated along the prc‐central gyrus with the width of l0mm and the length of 20mm. Watanabe et a1 reported rCBV increase in the pericentral region during fingcr movements using NIRS mapping. It was clearly shown that the rCBV inci‐ease was closely correlated with that of IHbTotal I and [HbOxyl, observed with NIRS. The activity‐coupled CBV increase is also expected during epileptic scizxres. Horsley first descrihed ictal focal hyperperfusion by the direct visual inspection of the brain during seizui‐es. This was confirmed by the PET obscrvation unexpcctedly taken during seizurcs. Ictal SPECT is the first well‐planned examination to reveal the ictal hyperpcrfilaion . Extratemporal epilepsy has become a major target for epilcptologists i n focus detection espccially when a structural lesion cannot be detected. In such conditions, intracranial EEG recording is essential. As the interictal hypoinetaholism is less frequently found in thcse cases, the ictal SPECT has hccn considered 21s major information to determine the optimal placement of intracranial electrodcs. Conclusion: Based on the present results, wc i.eached a conclusion that the combination of ictal SPECT and NIRS mapping is one of the most reliable techniques to visualize the hernodynamic changes during a seizure. The former offcrs static tomographic images and thc latter continuous measurement supplemenkiry to each other.