Ontology-Driven Toolset for Audio-Visual Stimuli Representation in EEG-Based BCI Research

Abstract

In the last decade, the recent advances in software and hardware facilitate the increase of interest in conducting experiments in the field of neurosciences, especially related to human-machine interaction. There are many mature and popular platforms leveraging experiments in this area including systems for representing the stimuli. However, these solutions often lack high-level adaptability to specific conditions, specific experiment setups, and third-party software and hardware, which may be involved in the experimental pipelines. This paper presents an adaptable solution based on ontology engineering that allows creating and tuning the EEG-based brain-computer interfaces. This solution relies on the ontology-driven SciVi visual analytics platform developed earlier. In the present work, we introduce new capabilities of SciVi, which enable organizing the pipeline for neuroscience-related experiments, including the representation of audio-visual stimuli, as well as retrieving, processing, and analyzing the EEG data. The distinctive feature of our approach is utilizing the ontological description of both the neural interface and processing tools used. This increases the semantic power of experiments, simplifies the reuse of pipeline parts between different experiments, and allows automatic distribution of data acquisition, storage, processing, and visualization on different computing nodes in the network to balance the computation load and to allow utilizing various hardware platforms, EEG devices, and stimuli controllers.

How to Improve Utilizing Neural Interface by Means of Ontology-Driven Scientific Visualization Tools

“…Main focus points of this paper are devoted to tackling the following problems in order to improve the applicability of scientific visualization tools of SciVi developed by our team previously [1,2,3,4] in BCI studies to adapt them to the specific of the different IoT infrastructure, models and datasets:…”

“…That pipeline was developed using the SciVi platform, an ontology-driven scientific visualization and visual analytics toolset [1]. Then we extended this pipeline to apply it for the task of audiovisual stimuli presentation for the neurophysiological studies [2,3]. However, there were two major technical issues with this pipeline implementation that we're going to address in this paper:…”

“…This separation also allows us to reuse "Montage Provider" node at the stage of configuring experimental setup. In our previous paper [2,3] we demonstrated a data monitoring pipeline (Fig. 4), but as of time of writing it had some drawbacks; in particular, visualization of the headsets' impedances was static in reference to electrode counts and their positions.…”

Scientific Visualization Tools to Improve Utilizing Neural Interface

“…Main focus points of this paper are devoted to tackling the following problems in order to improve the applicability of existing scientific visualization tools developed by our team previously [1,2,3,4] in BCI studies:…”

“…That pipeline was based on the SciVi platform, an ontology-driven scientific visualization and visual analytics toolset [1]. Then we extended this pipeline to apply it for the task of audiovisual stimuli presentation for the neurophysiological studies [2,3]. However, there were two major technical issues with this pipeline that we're gonna address in this paper:…”