Additively manufactured (AM) lattice structures are applied in high-value applications such as lightweight aerospace design and biomedical implants. However, uncertainties of the geometry of as-manufactured AM lattice structures results in uncertainties in the associated mechanical response. This research proposes a non-destructive digital-twin certification methodology that quantifies the functional response of individual strut elements (and associated statistical distributions) from x-ray micro-computed tomography (mu CT) data for as-manufactured AM lattice structures. This methodology may be algorithmically applied, as is required for the cost-effective certification of high-value lattice structures. The proposed methodology is demonstrated for a digital twin of over 2000 strut elements within a Ti-6AI-4V lattice fabricated with laser-based powder bed fusion. This digital twin allows various geometric or functional analyses to be performed, and in this case is demonstrated by acquiring statistical distributions of the predicted critical buckling load as a function of the strut element build orientation.