Figure 3. Two modalities for training nnU-Net. (A) The original 3D TOF-MRA, (B) the normalized and enhanced vessel image extracted from the Hessian matrix.
All published articles are preserved here permanently: