Grant: Catalyst Gr: Addressing Musculoskeletal Conditions/Arthritis in Cnds - Military

The face is a complex 3D structure that is important to function (i.e. sight, speech, chewing, and sense of smell) and an individual's self-perception (physical appearance). With advances in body armor, severe facial trauma now represents 30% of military war injuries. After a traumatic facial injury an individual's primary desire is to return to their original state, both in terms of function and their appearance. However, there is often insufficient information on the shape of an individual's pre-injury face and the underlying bone. This makes accurate restoration of appearance through surgery extremely challenging, with much of a plastic surgeon's work dependent on their experience and visualization abilities. In society today, we have many photographs of ourselves in frames and albums, and on our phones, digital cameras and social media sites. Technological innovation in the field of computer vision enables the generation of 3D models from 2D pictures, but the accuracy of these models to guide surgical facial reconstruction needs to be high. This project aims to develop a workflow to guide surgical reconstruction of the face by integrating and optimizing 2D to 3D technologies (in computer vision, manufacturing and forensics) that take advantage of pre-injury photos and the relationship between our appearance and the underlying bone structure. This project will optimize the ability of 3D face models to provide information that is accurate enough for surgical reconstructive planning, and validate the models using 3D imaging with respect to the face and underlying bone geometry. Physical surgical guides (for cartilage and bone) will be created and tested based on the modeled 3D face and bone structures. Ultimately these technologies will be integrated with our current skin template workflow into a highly automated surgical planning pipeline to improve facial restoration after severe trauma that often occurs in the military context.