Children born with congenital heart disease often have uniquely shaped hearts, and standard medical devices are rarely designed with a specific child’s anatomy in mind. To address this, Children’s Hospital of Philadelphia (CHOP) uses open source AI tools, including MONAI and SlicerHeart, to convert a child’s CT, MRI or ultrasound scans into detailed 3D heart models that help doctors understand each child’s anatomy before surgery. These models are already used routinely to plan surgeries for complex heart defects.
CHOP is also working with NVIDIA and the open-source community to use the Warp framework and Newton physics engine to simulate how devices may behave inside a child’s heart, while separately developing digital twin technology with NVIDIA Omniverse and OpenUSD. By combining open source imaging, simulation and visualization tools, this collaborative effort aims to give clinicians a clearer, more personalized understanding of each child’s heart before treatment decisions are made.
Image generated by: ChatGPT
