Abstracts

Rationale: To investigate seizure suppression efficacy using closed-loop stimulation at earlier and later stages of seizure formation.Methods: Acute and chronic seizures were induced in adult Wistar rats using 4-aminopyridine (15 rats: 5 in control and 10 in stimulation groups) and kainic acid (10 rats: 5 in control and 5 in stimulation groups), respectively. In the experiments, two bipolar microelectrodes were implanted in both hippocampi (CA1 area) and a real-time feedback stimulation paradigm was implemented which triggered a short periodic electrical stimulus (5 Hz for 5 sec) upon detecting a seizure precursor. This detection algorithm was based on continuous monitoring of real-time phase synchronization analysis in between two hippocampi. This paradigm was first implemented using a custom made microchip for 256-channel amplifier and 64-channel stimulator, and a laptop for phase synchronization analysis; which was called prototype CLS-V1. Later, the same paradigm was implemented in another custom made microchip for 64-channel amplifier, 64-channel stimulator and signal processor; which was called prototype CLS-V2. The CLS-V1 and CLS-V2 were used in the acute (a total of 10 rats, 5 for each prototype, and each experiment lasted for 2 hours) and chronic (five rats and each experiment lasted for 24 days: 6-day no-stimulation + 6-day stimulation using CLS-V1 + 6-day no-stimulation + 6-day stimulation using CLS-V2) experiments for seizure suppression.Results: Baseline mean seizure frequencies of the control groups were 13.2 seizures per hour (acute) and 5.39 seizures per day (chronic). CLS-V1 and CLS-V2 detected 84% and 91% of acute seizures with a mean false alarm of 0.47 and 0.25 per hour and 80% and 87% of chronic seizures with a mean false alarm of 0.67 and 0.52 per day, respectively. The CLS-V1 detected seizures on average 21 and 53 seconds before a seizure onset (signal amplitude greater than twice the standard deviation of background activities) in acute and chronic models, respectively; however, CLS-V2 detected seizures on average 2 and 4 seconds after the seizure onset in acute and chronic models, respectively. CLS-V1 reduced the seizure frequencies to 2.4 seizures per hour (81% suppression in acute) and 0.41 seizures per day (90% suppression in chronic). On the other side, CLS-V2 reduced the seizure frequencies to 5.54 seizures per hour (58% suppression in acute) and 1.29 seizures per day (76% suppression in chronic).Conclusions: A brief closed-loop stimulation at earlier stage of seizure formation is more effective to stop the seizure development than a later stage stimulation, presumably because it is easier to perturb the incipient synchronization before the full paroxysm develops than stopping the highly synchronous activities during the paroxysm. Abnormal synchrony patterns are common features in other neuropsychiatric syndromes, thus, this type of closed-loop stimulation paradigm could be a novel therapeutic modality for various mental disorders.