Given　the　complex　temporal　evolution　of　epileptic　seizures,　understanding　their　dynamic　nature　might　be　beneficial　for　clinical　diagnosis　and　treatment.　Yet,　the　mechanisms　behind,　for　instance,　the　onset　of　seizures　are　still　unknown.　According　to　an　existing　classification,　two　basic　types　of　dynamic　onset　patterns　plus　a　number　of　more　complex　onset　waveforms　can　be　distinguished.　Here,　we　introduce　a　basic　three-variable　model　with　two　time　scales　to　study　potential　mechanisms　of　spontaneous　seizure　onset.　We　expand　the　model　to　demonstrate　how　coupling　of　oscillators　leads　to　more　complex　seizure　onset　waveforms.　Finally,　we　test　the　response　to　pulse　perturbation　as　a　potential　biomarker　of　interictal　changes.