Noninvasive Seizure Screening in Preclinical Models of Epilepsy There is an urgent need for research into treatment options for epilepsy and other seizure disorders. Animal models are increasingly used to understand disease mechanisms and to screen promising therapeutic approaches. Animal epilepsy model use typically requires expensive and labor-intensive experimentation, with invasive EEG measurements being the preferred method of validation. This severely limits the pace and scale of investigation. In models of acquired epilepsy, animals usually undergo treatment to induce status epilepticus? a period of unremitting seizure?followed by a latent period during which the brain rewires itself to generate spontaneously recurring seizures, evidence of chronic epilepsy. The duration of the latent period and the likelihood that an animal will develop epilepsy are both uncertain. Animals must be observed for weeks to confirm epilepsy before they are ready for experimentation. During this latent period, seizures are commonly documented by visual observation or video review, which are tedious and prone to error. A commercial system for automated noninvasive seizure detection would therefore be attractive to epilepsy researchers. Signal Solutions, LLC, has developed technology based on piezoelectric sensors for noninvasive, high- throughput behavioral monitoring of rodents that is currently used by research groups around the world to identify genes related to sleep and circadian rhythms. The system discriminates sleep from wakefulness with over 90% accuracy. The Sunderam Lab at the University of Kentucky has further demonstrated using EEG analysis that these piezo sensors can be used to label REM and NREM stages of sleep in mice. In this STTR proposal, PI Sunderam and co-investigator Bauer will work with Signal Solutions to develop methods based on the piezo technology for accurate noninvasive seizure screening in rodent models of epilepsy. The result will be a validated system that minimizes the need for invasive and resource-intensive EEG analysis or tedious video monitoring. The objectives of the project are the following: 1. Test the feasibility of coarse detection of epilepsy onset in rodents using a noninvasive piezo sensor; 2. Train a piezo classifier to accurately detect and quantify seizures in rodents with confirmed epilepsy; and 3. Test the utility of a miniature piezo sensor for seizure screening and an infrared imager for seizure verification and severity assessment. The envisioned product is a turnkey system for convenient and noninvasive seizure screening in small animal models of epilepsy in custom or commercial cages. Potential customers include academic research labs as well as labs in the pharmaceutical sector engaged in high-volume screening of antiepileptic drugs.
The development of treatments for epilepsy depends heavily on laboratory animal studies in which the effect of treatment on seizures is measured. Seizures are rare and unpredictable, and monitoring animals for seizures requires tedious observation or review of video recordings; the alternative is to record brain signals, which requires invasive surgery to implant electrodes. A new technology is proposed here that would enable completely noninvasive detection of seizures in animals. The accuracy of this method will be compared with EEG analysis in an animal model of epilepsy.
