EBAPy: A Python framework for analyzing the factors that have an influence in the performance of EEG-based applications

Dustin Carrión-Ojeda ,Paola Martínez-Arias,Israel Pineda,Rigoberto Fonseca-Delgado

doi:10.1016/j.simpa.2021.100062

Dustin Carrión-Ojeda , Paola Martínez-Arias + Show 2 more

Open Access

https://doi.org/10.1016/j.simpa.2021.100062

Copy DOI

Abstract

EBAPy is an easy-to-use Python framework intended to help in the development of EEG-based applications. It allows performing an in-depth analysis of factors that influence the performance of the system and its computational cost. These factors include recording time, decomposition level of Discrete Wavelet Transform, and classification algorithm. The ease-of-use and flexibility of the presented framework have allowed reducing the development time and evaluating new ideas in developing biometric systems using EEGs. Furthermore, different applications that classify EEG signals can use EBAPy because of the generality of its functions. These new applications will impact human–computer interaction in the near future.

Highlights

Most bioelectrical signals are traditionally used in medical diagnosis
In recent years, researchers have focused on using these signals in different applications such as biometric systems [1]
The decomposition level of this transform can be a factor that affects the performance of the system

Summary

Introduction

Most bioelectrical signals are traditionally used in medical diagnosis. in recent years, researchers have focused on using these signals in different applications such as biometric systems [1]. This level has a direct impact on the overall computational cost For this reason, EEG-Based Applications using Python (EBAPy) enables conducting an analysis to select the best DWT decomposition level for a specific application. EEG-Based Applications using Python (EBAPy) enables conducting an analysis to select the best DWT decomposition level for a specific application It allows an assessment of the recording time required to achieve a high system performance. It selects the best classification algorithm for the signals among five classifiers: The code (and data) in this article has been certified as Reproducible by Code Ocean: (https://codeocean.com/). Support Vector Machine (SVM), K-nearest Neighbors (KNN), AdaBoost (AB), Random Forest (RF), and Multilayer Perceptron (MLP)

Impact overview

Expected impact

Software functionality

Conclusions and further development