Human-Computer Interfaces (HCI) are increasingly ubiquitous in multiple applications including industrial design, education, art or entertainment. As such, HCI could be used by very different users, with very different skills and needs. This thus requires user-centered design approaches and appropriate evaluation methods to maximize User eXperience (UX). Existing evaluation methods include behavioral studies, testbeds, questionnaires and inquiries, among others. While useful, such methods suffer from several limitations as they can be either ambiguous, lack real-time recordings, or disrupt the interaction. Neuroergonomics can be an adequate tool to complement traditional evaluation methods. Notably, Electroencephalography (EEG)-based evaluation of UX has the potential to address the limitations above, by providing objective, real-time and non-disruptive metrics of the ergonomics quality of a given HCI (Frey 2014). In this abstract, we present an overview of our recent works in that direction. In particular, we show how we can process EEG signals in order to derive metrics characterizing 1) how the user perceive the HCI display (HCI output) and 2) how the user interacts with the HCI (HCI input).