romainvala / r_qc Goto Github PK

edit init_path.R and change the path, to your current path


%first run the histogram fitting with the R command : 


cd dir where the data.nii.gz  is 
 Rscript /fullpath /volume_hist_bg_create_mask.R data.nii.gz


If there is a file name bg_mask_crop.nii.gz in the current dir, it will take the background mask to compute signal std, if not it will create the background mask (For the paper I took your background mask or a version eddited manually (not sure)



this will create a csv with std in backgroup and the different peak



x_pos_sig : will be the mean peak Signal level (which depend of the total number of peak found as follow

    if x_data_P4_pos(k)>0
        x_pos_sig(k) = (cout.x_data_P4_pos(k)+cout.x_data_P3_pos(k)+cout.x_data_P2_pos(k))/3;
    elseif x_data_P3_pos(k)>0
        x_pos_sig(k) = (cout.x_data_P3_pos(k)+cout.x_data_P2_pos(k))/2;

    elseif x_data_P2_pos(k)>0
        x_pos_sig(k) = cout.x_data_P2_pos(k);
    else
        x_pos_sig(k) = cout.x_data_P1_pos(k);
    end
 

then SNR is 
	x_pos_sig/noise_std

(noise_std is the field in the csv file)