Activation outside the brain

I am doing an analysis using DPRSFA V_2.1 and Matlab 2015b.​ ​Unfortunately, when I complete the below-described analysis, ​I ​am getting​ ​significant ​activation outside of the brain​ (please see ​the picture in ​the ​attachment​)​. ​Would anyone know what could be the cause of th​is​? 
​To describe my analysis, ​I have data from two groups of subjects, acquired with different scanning parameters. One group: TR 2.3, vox size 3x3x3, second group TR 2.3 vox size 4x4x4.
Steps I did:
Preprocessing (separately for both groups):
Remove first 10 time points
Segmentation (New Segment + DARTEL) 
Normalisation (by DARTEL; Voxel Size: 4x4x4)
Nuisance Covariates Regression (Rigid Body 6, method: CompCor)
Smoothing (6 6 6)
Functional Connectivity (with defined ROIs - spheres)
Second Level Analysis (for both groups together) - in SPM12. 
I would really appreciate any help in identifying the problem.
Best regards,
Emilia Cheladze
Image icon ACC.jpg58.94 KB

The outside brain results might be caused by coverage issues. You can ignore them.