plot_speech_strf.m

function plot_speech_strf
%PLOT_SPEECH_STRF  Plot example speech STRF.
%   PLOT_SPEECH_STRF loads an example dataset, estimates and plots a speech
%   STRF and the global field power (GFP) from 2 minutes of 128-channel EEG
%   data as per Di Liberto et al. (2015).
%
%   Example data is loaded from SPEECH_DATA.MAT and consists of the
%   following variables:
%       'stim'      a vector containing the speech spectrogram, obtained by
%                   band-pass filtering the speech signal into 128
%                   logarithmically spaced frequency bands between 100
%                   and 4000Hz, taking the Hilbert transform at each band
%                   and averaging over every 8 neighbouring bands.
%       'resp'      a matrix containing 2 minutes of 128-channel EEG data
%                   filtered between 0.5 and 15 Hertz
%       'fs'        the sample rate of STIM and RESP (128Hz)
%       'factor'    the BioSemi EEG normalization factor for computing the
%                   STRF in microvolts (524.288mV / 2^24bits)
%
%   mTRF-Toolbox https://github.com/mickcrosse/mTRF-Toolbox

%   References:
%      [1] Di Liberto GM, O'Sullivan JA, Lalor EC (2015) Low-Frequency
%          Cortical Entrainment to Speech Reflects Phoneme-Level
%          Processing. Curr Biol 25:1-9.
%      [2] Crosse MC, Di Liberto GM, Bednar A, Lalor EC (2016) The
%          multivariate temporal response function (mTRF) toolbox: a MATLAB
%          toolbox for relating neural signals to continuous stimuli. Front
%          Hum Neurosci 10:604.
%      [3] Crosse MJ, Zuk NJ, Di Liberto GM, Nidiffer AR, Molholm S, Lalor
%          EC (2021) Linear Modeling of Neurophysiological Responses to
%          Speech and Other Continuous Stimuli: Methodological
%          Considerations for Applied Research. Front Neurosci 15:705621.

%   Authors: Mick Crosse <mickcrosse@gmail.com>
%   Copyright 2014-2020 Lalor Lab, Trinity College Dublin.

% Load data
load('../data/speech_data.mat','stim','resp','fs','factor');

% Normalize data
resp = factor*resp;

% Model hyperparameters
Dir = 1; % direction of causality
tmin = -100; % minimum time lag
tmax = 400; % maximum time lag
lambda = 0.5; % regularization value

% Compute model weights
% Note, ridge regression is used instead of Tikhonov regularization to
% avoid cross-channel leakage of the multivariate input features
model = mTRFtrain(stim,resp,fs,Dir,tmin,tmax,lambda,'method','ridge',...
    'split',5,'zeropad',0);

% Plot STRF
figure('Name','Speech STRF & GFP','NumberTitle','off')
set(gcf,'color','w')
subplot(1,2,1)
mTRFplot(model,'mtrf','all',85,[-50,350]);
title('Speech STRF (Fz)')
ylabel('Frequency band')

% Plot GFP
subplot(1,2,2)
mTRFplot(model,'mgfp','all','all',[-50,350]);
title('Global Field Power')