Epilepsy is a pathological condition characterized by excessive brain electrical activity, resulting in sudden seizures perceived by patients, which can lead to sudden loss of consciousness and convulsions. Diagnosing epilepsy requires the physician to manually inspect the electroencephalography of the patient, a typically time-consuming task for medical staff. In this paper, we propose a method for the potential identification of epileptic seizures, by generating from an electroencephalography two different types of wavelet transforms i.e., the Morlet and the Mexican Hat, obtained by previously transforming the waveform into an audio file. We exploit a hybrid quantum-classical model for the binary classification of an electroencephalography as belonging to an epileptic patient or a not epileptic one, directly comparing the proposed model with state-of-the-art convolutional neural networks, in order to demonstrate the effectiveness of the proposed method, obtaining that the proposed hybrid model is able to obtain better performances from the accuracy point of view with respect to convolutional neural networks.