ScienceShot: Making Virtual Humans Dance
MIRALab - University of Geneva

Click here for free access to our latest coronavirus/COVID-19 research, commentary, and news.

Support nonprofit science journalism

Science’s extensive COVID-19 coverage is free to all readers. To support our nonprofit science journalism, please make a tax-deductible gift today.

ScienceShot: Making Virtual Humans Dance

CHICAGO, ILLINOIS—The motion capture technology that sent Na'vi prancing through Pandora with lifelike accuracy in the film Avatar has another use in the lab: creating fine-scale, highly personalized models of how a body moves. Computer scientist Nadia Magnenat Thalmann of the University of Geneva in Switzerland and colleagues have been working to design virtual replicas of athletes’ bodies—not just their general shape and movement, but also dynamics hidden within the joints that bear the strain. Their first test subjects: a group of professional ballerinas from the Grand Theater of Geneva. Due to the unusual stress on their joints as they dance, many ballerinas need hip replacement surgery in their early teens, says Thalmann, who described the ongoing modeling project here today at the annual meeting of AAAS, which publishes Science. The case study used MRI to generate a model of each dancer’s muscles, cartilage, and bone, slice by slice. Then, by adding data from a motion sensor suit (like the one in the photo above), the researchers watched how stress was distributed within the tissues (colored joint models, left) as she performed. The results have helped each dancer understand which movements put her at risk of long-term damage, Thalmann says, and predict the need for future surgery. This technology is still young, Thalmann points out, and the process of analyzing personalized data takes about a month. But she predicts that doctors will someday use the method to quickly model strain and injury in the clinic … even for those who don’t regularly subject their bodies to high kicks and splits.

See more of our coverage from AAAS 2014.

See more ScienceShots.