The objective of my neuroscience research was to analyse trends in brain EEG activity during affective picture viewing.
I used MATLAB signal-processing and statistical toolbox to compute features (alpha-beta asymmetry ratio, alpha power, beta power, theta power) of brain EEG signals across 32 channels and five subjects and analyzed trends in these features in relation to subject's valence/arousal ratings after viewing affective IAPS images
Please refer to the files below to see my project results:
-
Read my research paper (Note: pdf file might take a while to load)
-
View my research poster:
loadeegdata.m – given a subject name, reads the EEG file, and generates details raw EEG data in matlab matrices
extract.m – returns a segment of signals from raw EEG data, given starting time, signal length, user’s raw EEG data, and channel number, and returns a vector of time and a vector of amplitudes
generalFilter.m – returns a filtered signal after detrending and filtering a raw signal given highpass and lowpass value
loadPxxSubj.m - calculates features of brain signals (alpha band, beta band, theta band) using high-pass and low-pass filters and pwelch.
loadPxxSubjPrint.m – plots the pwelch of a frequency band for high, medium and low levels of valence/arousal for one subject
getbaRatio.m – returns the beta-alpha ratio given a signal
getbaRatioWBase.m – returns the beta-alpha ratio given a signal and a baseline
findSubjRatings.m – returns a list of pictures that the subject evaluated, the valence/arousal rating of the subject, and additionally, the valence/arousal rating from IAPS, given subject based on file list provided by loadRatings.m
loadRatings.m - returns a list of files, reads the IAPS manual, and tabulates the valence/arousal ratings of all IAPS pictures
barAlphaAsymmetry.m – compares alpha asymmetry or theta asymmetry differences of 32 channels for five subjects for each valence/arousal class
barBandPower.m – compares signal powers (alpha, beta, theta) or beta-alpha ratio differences of 32 channels for five subjects for each valence/arousal class
barPower.m - compares average signal powers (alpha, beta, theta) or beta-alpha ratio of three categories of valence/arousal for 32 channels for five subjects for each valence/arousal class
trendAlphaAsym.m – plots a scatter graph of alpha asymmetry or theta asymmetry against valence/arousal rating for 1 subject
trendBandPower.m – plots a scatter graph of average beta power or alpha power or theta power or beta-alpha ratio against valence/arousal rating for 1 subject`
loadSignals.m – plots the average segment of EEG signals of a certain frequency band averaged across all subjects in high, medium and low levels of valence/arousal