Oscillatory episodes of brain activity derived from the OMEGA dataset

This dataset contains source-reconstructed magnetoencephalography (MEG) brain activity (1,925 voxels) from a sample of 128 healthy participants from The Open MEG Archive (OMEGA, version 2.0; Niso et al., 2016). For each participant, episodes of oscillatory brain activity were identified using sBOSC (Stern et al., 2025; github.com/necog-UAM/sBOSC.-Oscillation-Detection-Algorithm-), an extension of the BOSC (Better OSCillation Detection) family of algorithms.

As in previous BOSC-based approaches, sBOSC detects oscillatory episodes that exceed both a predefined power threshold and a minimum duration criterion. In addition, sBOSC incorporates spatial information into the detection procedure by identifying local maxima across the reconstructed brain volume. Oscillatory episodes are further constrained to frequencies corresponding to peaks in the power spectra, ensuring that the detected events reflect genuine oscillatory activity.