Decoding Imagined Musical Pitch from Human Scalp Electroencephalograms

Fea_total is a feature set extracted from all EEG channels. Fea_diff is a feature set extracted from EEG channels considering the lateral differences between two hemispheres.

Each of Fea_total and Fea_diff has 20 cells corresponding to subject numbers, and each cell has all/selected corresponding to all feature dimensions or selected feature dimensions, and those are divide to train/test.

Therefore, if you want to access the train data of subject1's feature extracted from all channels but features after dimension selection, you can enter Fea_all{1}.selected.train.

labeltr is label if train set and labete is label of test set. They have label from 1 to 7, and those are corresponding to C4 to B4.