--- Overview ---

This auditory repository has been converted from BIDS (Brain Imaging Data Structure) to CND (Continuous-event Neural Data).

Please cite the original paper from Prof. Tom Francart's team:
SparrKULee: A Speech-evoked Auditory Response Repository of the KU Leuven, containing EEG of 85 participants
Bernd Accou, Lies Bollens, Marlies Gillis, Wendy Verheijen, Hugo Van hamme, Tom Francart
bioRxiv 2023.07.24.550310; doi: https://doi.org/10.1101/2023.07.24.550310
Now published in Data doi: 10.3390/data9080094


BIDS-to-CND conversion by Cillian French and Giovanni Di Liberto, Trinity College Dublin.
Di Liberto, Nidiffer, et al., “A standardised open science framework for sharing and re-analysing of neural data acquired to continuous sensory stimuli”. arXiv, 2023, https://arxiv.org/abs/2309.07671.

The structure and use of this dataset is validated through the undergraduate capstone project "Facilitating Collaboration in Cognitive Neuroscience Through EEG Data Reformatting and Analysis" by Cillian French (supervised by Prof. Giovanni Di Liberto).


--- SparrKULee in CND ---

The CND version of the SparrKULee dataset is of 78 subjects with trials split into 2 datasets based on stimulus condition used.

SparrKULee 1 involves audiobook listening.

SparrKULee 2 involves podcast listening.

This dataset is available in two versions: Full and Lite (see below)


--- Full version ---

This version is of original sampling size (1024Hz) and no preprocessing steps have been applied.

!!! WARNING !!! - The stimuli envelopes are of 64Hz sampling rate which is different to that of the EEG data sampling rate, making them unsuitable for direct comparison before preprocessing.


--- Lite version ---

This version is preprocessed for size purposes with the following steps applied:

- Low-pass filter of 8Hz (zero-phase shift Butterworth filter with order 2).
- Downsampled to 64Hz (matching the sampling frequency of the stimulus features).