[Freesurfer] lh.ROI45.label and lh.ROI44.label overlapping

Hi freesurfers experts, this is (please check attached image) the best I got when trying to map a volume ROI (BA45 in MarsBar = MNI_Frontal_Inf_Tri_L_roi) to freesurfer using mri_vol2surf. I used /freesurfer/average/mni152.register.dat instead of spmregister since the ROI was already in mni152 2mm, the registration was perfect -checked with tkregister2.

- I obtained the red area of the image, which is outside of the ROI45.label. Considering that the source was not exactly the same, could I consider it acceptable for cortical thickness analysis? - Why are the 44 and 45 overlapping?

Thank you very much and sorry for the multiple questions, Gari