This paper presents summary of a paper review on EEG signals focusing on signal acquisition design. The discussion is devoted to some prominent research on imagined speech recognition, which is divided into vowel recognition, syllable recognition and word recognition. This explanation provides an overview state of the art from acquisition design to imagined speech recognition research and there is an open opportunity to develop an EEG signal classification model for recognizing some imagined speech with better accuracy and faster recognition process. Hopefully, this presentation can be considered as a reference for new researchers in the field of BCI in designing an acqusition
protocol that suits their purpose.