Quantitative EEG features and machine learning classifiers for eye-blink artifact detection: A comparative study

A Hardware-Based Configurable Algorithm for Eye Blink Signal Detection Using a Single-Channel BCI Headset

“…Therefore, studies focus on the waves generated in the electrodes in these positions. Although the signals returned by the EEG are raw signals, the processing is usually performed with some type of feature derived from these signals [1,2]. The most significant examples of these features are as follows:…”

“…The first motivation involves the elimination and/or adaptation of electroencephalographic (EEG) signals, since the presence of artifacts renders such signals unusable for drawing conclusions (e.g., diagnoses). In the former case, the data set that needs to be processed is reduced, and in the latter case, the quality of the data set to be processed is improved [1][2][3]. These studies typically include healthy individuals as well as those with some kind of pathology.…”

“…The system to detect this artifact must have a differentiator to eliminate the DC value, and a low-pass filter to eliminate the higher-frequency wave due to brain activity. Equation (1) shows the implementation of the differentiator as an approximation of its definition.…”