Best Practices in Data Analysis and Sharing in Neuroimaging using MEEG

Abstract

Neuroimaging methods, including magnetoencephalography and electroencephalography (MEEG), allow non-invasive collection of neural data in healthy people and in individuals with neurological or psychiatric disorders, with the aim of advancing the understanding of brain function in health and disease. Currently, scientific practice is undergoing a tremendous change, aiming to improve both research reproducibility and transparency in data collection, documentation and analysis, and in manuscript review. To advance the practice of open science, the Organization for Human Brain Mapping originally created the Committee on Best Practice in Data Analysis and Sharing (COBIDAS), which produced a report for MRI-based data. This effort continues with the OHBM's creation of a COBIDAS MEEG committee whose task is to create a similar document that describes best practice recommendations for MEEG data. The document has been drafted by OHBM experts in MEEG, with input from the brain imaging community including OHBM members who volunteered to help with this effort and executive committee members of the International Federation for Clinical Neurophysiology. This document outlines the principles of open and reproducible research in MEEG, and also applies these principles to specific research practices. Not all MEEG data practice is described in this document. Instead, we present information that we believe should be reported to fully understand and potentially replicate any study. We describe both good practice and reporting standards that ensure data collection and analysis will adhere to the highest standards. We also deal with issues of nomenclature and terminology which, due to the relatively long history of MEEG, have come about as different subfields emerged.

How Sensitive are EEG Results to Preprocessing Methods: A Benchmarking Study

“…We then computed correlations of the corresponding fingerprint vectors associated with pairs of preprocessing methods to summarize how much preprocessing affects spectral results. In addition, we averaged each spectrogram within standard frequency bands (delta: [2,4] Hz, theta: [4,7] Hz, alpha: [7,12] Hz, beta: [12,30] Hz) to form separate fingerprints for each band and computed correlations across corresponding fingerprint bands for pairs of preprocessing methods.…”

“…The disagreement between the ICA-based methods (LARG and MARA) and the ASR-based methods (ASR_10* and ASR_5*) in the delta frequency band ( [2,4] Hz for this analysis) is likely due to the differences in baselining and highpass filtering that occurred at the beginning of the respective pipelines. However, ASR_10* and ASR_5* used the same input signals and even in this case, the correlations in the delta bands were much lower than in other bands.…”

Combined behavioral and electrophysiological evidence for a direct cortical effect of prefrontal tDCS on disorders of consciousness

“…All statistical analyses were performed with python using scipy (78), mne-python (61) and scikit-learn (63) packages, except for the non-parametric factorial analysis performed in R statistical software with the ARTool (79) package and electric fields analysis performed in Matlab software. Throughout all EEG analyses, from preprocessing to statistical analyses, we tried to comply with the CODIBAS-MEEG Best Practices in Data Analysis and Sharing in Neuroimaging using MEEG initiative (80).…”

BIDS-EEG: an extension to the Brain Imaging Data Structure (BIDS) Specification for electroencephalography

“…The long history, versatility and variety of applications of EEG makes it a data and method rich technique. The recent OHBM guideline for good practices and reproducibility in EEG (Pernet et al, 2018) lists eight preprocessing steps for standard event related potentials (identification and removal of electrodes with poor signal quality, artifact identification and removal, detrending, digital low-and high-pass filtering, data segmentation, additional identification/elimination of artifacts, baseline correction, and rereferencing) with the order of steps depending on applications and potentially augmented by additional transformation in time, frequency or time-frequency domains, projection into source space and additional connectivity measurements. This implies that while the BIDS-EEG will help data sharing, it will remain non-trivial to develop automated preprocessing pipelines of magneto- Niso et al (2018) and electrophysiological (Holdgraf et al, in preparation) prepared data such as fMRI BIDS apps (Gorgolewski et al, 2017).…”