Our trial includes 20 CFS patietns and 20 healthy controls.
This is the 400 ROIs information of Schaefer template provided by DPABISurf.
This is the information about 60 features we selected from the 160000 ROIs.
This is the functional connectivity values of the 60 features, which has been processed by fisher Z translation. The order of this 80 columns is: 20 patients before intervention, 20 healthy people before intervention, 20 patients after intervention, 20 healthy people after intervention.
This is the final model built by RandomForest Classification.
This is the original dataset (have been fisher Z translation) before intervention, columns 1-20 are CFS patients, columns 21-40 are healthy controls. Each subject has 160000 values of functional connection via Shaefer400 template.
This is the original dataset (have been fisher Z translation) after intervention, columns 1-20 are CFS patients, columns 21-40 are healthy controls. Each subject has 160000 values of functional connection via Shaefer400 template.
This can be performed in python3 to test our the predictive ability of our final model.
This is the train dataset of 60 features generated by the shuffle split, the split ratio is 0.5.
This is the test dataset of 60 features generated by the shuffle split, the split ratio is 0.5.
The x_train and x_test are consisted of functional connections of 60 features of subjects before the intervention (20 patients and 20 healthy people).