MedForj: An open, large-scale foundational generative prior for high-resolution 3D brain MRI

Excerpt

arXiv:2510.26834v2 Announce Type: replace-cross Abstract: This work introduces MedForj, a suite of 3D foundational generative priors based on diffusion models. The MedForj models were trained on $72{,}659$ 1~mm isotropic 3D $T_1$-weighted MRI human brain image volumes from $38{,}174$ subjects, drawn from a curated corpus of $80{,}675$ volumes from $42{,}506$ subjects spanning $38$ publicly available datasets. These training images were manually inspected to exclude those with poor quality and ex