Robin’s Viewer: Using deep-learning predictions to assist EEG annotation

Abstract

Machine learning techniques such as deep learning have been increasingly used to assist EEG annotation, by automating artifact recognition, sleep staging, and seizure detection. In lack of automation, the annotation process is prone to bias, even for trained annotators. On the other hand, completely automated processes do not offer the users the opportunity to inspect the models’ output and re-evaluate potential false predictions. As a first step toward addressing these challenges, we developed Robin’s Viewer (RV), a Python-based EEG viewer for annotating time-series EEG data. The key feature distinguishing RV from existing EEG viewers is the visualization of output predictions of deep-learning models trained to recognize patterns in EEG data. RV was developed on top of the plotting library Plotly, the app-building framework Dash, and the popular M/EEG analysis toolbox MNE. It is an open-source, platform-independent, interactive web application, which supports common EEG-file formats to facilitate easy integration with other EEG toolboxes. RV includes common features of other EEG viewers, e.g., a view-slider, tools for marking bad channels and transient artifacts, and customizable preprocessing. Altogether, RV is an EEG viewer that combines the predictive power of deep-learning models and the knowledge of scientists and clinicians to optimize EEG annotation. With the training of new deep-learning models, RV could be developed to detect clinical patterns other than artifacts, for example sleep stages and EEG abnormalities.

Artificial Intelligence in Newborn Medicine

“…Plotly can also be used in conjunction with Dash for building interactive web applications. 145,146 • ggplot2 is a plotting system for the R programming language inspired by the Grammar of Graphics. 147 It is a powerful framework for creating statistical graphics such as bar plots, pie charts, and histograms.…”

SLAM – a Thin-Client for Interoperable Annotation and Biomedical Signal Handling

“…Inspired by seminal contributions ten years ago, especially the popular EEG analysis toolbox MNE [16], some efforts were geared towards developing alternative software tools dedicated to this task. One can cite, non-exhaustively, Robin Viewer [17], an open-source, platform-independent, interactive web application, designed for allowing deeplearning paradigms; a clinical EEG research platform [18] designed to allow progressive model construction, including in particular EEG visualization and annotation; copla-editor [19], a web-based visualization, assessment and collaboration tool for multidimensional biosignals; and the time series plug-in of CrowdCurio [20], a generic annotation library with support for EEG data.…”

“…While CrowdCurio [20] claims to support EEG annotation, it is unclear how to run the program [10]. Regarding the other two freely available software platforms, namely Robin's Viewer [17] and copla-editor [19], our contribution shares common properties and purposes, but also differs in various points. Both projects implement an EEG viewer, with the possibility to add annotations in a web-based user interface.…”

“…Both projects implement an EEG viewer, with the possibility to add annotations in a web-based user interface. [17] makes extensive use of the MNE library, which enables signal processing tools such as band-pass filters. It can also be used for active learning of a segmentation task, such as artifact detection.…”

“…However, these projects do not work solely on a web browser, and do not provide a solution for interoperable annotations. We also note that [17] and [19] must deal with EDF+-compatible annotations, which are single-channel, ambiguous, and do not allow global descriptive meta-data. Other approaches require to keep data and meta-data in the same file, such as BIDS-EEG or SEDS [21].…”