How it works

The model ingests 3D skeletal motion captured during walking and learns generalizable representations. Those representations are then mapped to health phenotypes across organ systems.