The performance of biomaterial scaffolds for bone tissue engineering, as porous titanium implants, is strongly dependent of its structural features. A reliable structural characterization of this kind of implant is very important. The most of image analysis techniques just supplies 2D information about the structure of specimens. X-ray microtomography is imaging technique that can produce 3D images of samples, however, stochastic models can also estimate properties of porous materials in 3D. This work presents the evaluation of a 3D model (using a truncated Gaussian method) in comparison to 3D microtomography volume, both from a titanium scaffold sample. In order to compare, geometrical parameters were measured for both 3D volumes. By the results, the truncated Gaussian 3D method reproduced a model with similar values to the microtomography volume, showing a good agreement among data, which suggests the use of this technique to estimate physical parameters of titanium scaffolds