Wrong ROI definition when extracting CSF/WM signal etc

Hello, I am receiving a bug when extracting the ROI signals (i.e white matter, CSF) for the nuisance covariates regression. The error appears after slice timing, realignment have been done. The error I received is below, I believe it is due to DPARSFA not being able to find the masks in the working directory. The error only appears with DPARSFA. My data was converted from DICOM --> NIFTI outside of DPARSF, then converted into spm format using freesurfers mri_convert to 208 indiviudal img files.