PubMed Health⌕ Search

Biomedical subjects

Dan Berg

Publications and source records attributed to Dan Berg.

2 recordsLinked to original sources

Real-time finite element based virtual tissue cutting.

Tool-tissue interaction is most accurately modeled as an imposed displacement constraint, thus augmenting the traditional finite element equation, Ku=f, to a 2 x 2 block system. This augmentation does two things: it enlarges the system, and it introduces the Schur complement(S) during the solution. This research has focused on efficient methods to update the Schur complement and it's inverse during displacement constraint removal and addition to allow for soft-tissue cutting to occur in real-time. By taking advantage of how the constraints impact the Schur complement, removal and addition of these constraints are handled by either performing a rank-2 or rank-1 update on S(-1), respectively. This greatly reduces the computational load on the CPU; and through use of timing tests shows that the update frequencies are well within an acceptable range for tactile feedback. To further solidify this method as being highly useful, a prototypical simulator has been developed that allows users to haptically perform bi-manual soft-tissue cutting.

Connective Tissue↗

Virtual image grafting: image based generation and visualization of virtual skin defects.

We have developed methods for rapidly generating 3d dermatologic datasets for use in education, training simulations and procedure planning. By compositing local surface features of cutaneous wounds onto patient images and 3d models, one can flexibly generate large patient variability from a small initial 3d library. The wound database is generated from clinically captured photographic images. The wound image is extracted from the image of the surrounding tissue. The 3d anatomy database is generated from MRI scans and from multiple photographic views. Extracted wound images can be moved rotated and scaled and blended in the final composite.

Databases as Topic↗