Alpha oscillations support the efficiency of guided visual search by inhibiting both target and distractor features in early visual cortex

The repository contains all scripts associated with a Magnetoencephalography (MEG) - Rapid Invisible Frequency Tagging study aimed at investigating the neural correlates of Guided Search.

authors: Katharina Duecker, Kimron L. Shapiro, Simon Hanslmayr, Jeremy Wolfe, Yali Pan, and Ole Jensen

MNE scripts

Maxfilter

Filtering based on maxwell equations and denoising of the sensors is implemented in MNE python.

Experiment code

experiment/a_exp is the main experiment file; only works with Propixx Lite projector and has been developed for MATLAB/2017a

MATLAB scripts

Used for all analyses in the manuscript

alpha

• a: TFR 4 - 30 Hz for each participant
• b: find Individual Alpha Frequency and sensors of interest
• c: align TFR to IAF
• d: alpha power per condition; performance ~ median split high/low alpha
• e: confirmatory analysis TFR for fast vs slow trials

behaviour

• a: behaviour per condition
• b: behaviour for alpha high/low
• c: confirmatory analysis: correlation reaction time ~ alpha power (mentioned in manuscript but not explicitly plotted)

cbrewer

brewer colormap

coherence

• a: coherence collapsed over trials for topo plots
• b: coherence per condition (priority map)
• c: coherence for fast vs slow (only weak relationship)
• d: coherence for alpha high vs low

eye movement

• a: ocular artefacts for alpha high vs low
• b: ocular artefacts for fast vs slow trials

preprocessing MEG

• a: merge .fif; .edf; and .mat (behaviour) files
• b: semi-automatic artefact rejection
• c: find the delay between trigger and diode onset (important to replace diode signal with sine wave)
• d: identify saccades (this is exploratory, there were too many saccades so we controlled for them with analyses mentioned above)
• e: ICA
• f: find tagging sensors of interest -> decide which stimuli to keep
• g: separate data into conditions (this saves one .mat files with MEG data and performance for each condition)

source: DICS beamformer

• a: align T1 to digitized fiducials
• b: estimate leadfield
• c: DICS for RIFT, alpha pre-search and during search
• d: plot