Currently, UBO Detector generates both image and text outputs (volumes and the number of WMH clusters) for whole brain, periventricular, deep, and lobar WMH, as well as WMH in arterial territories. This functionality is very useful, but I am interested in performing a more granular analysis by applying different masks, such as the Desikan atlas or aseg+aparc.
Could you provide guidance on how to apply these alternative masks for WMH segmentation and analysis?
Hello,
Currently, UBO Detector generates both image and text outputs (volumes and the number of WMH clusters) for whole brain, periventricular, deep, and lobar WMH, as well as WMH in arterial territories. This functionality is very useful, but I am interested in performing a more granular analysis by applying different masks, such as the Desikan atlas or aseg+aparc.
Could you provide guidance on how to apply these alternative masks for WMH segmentation and analysis?
Thank you.