NVIDIA Open-Source AI Cuts Pediatric Heart Modeling Time From Hours to Seconds

Children's Hospital of Philadelphia has deployed MONAI, an open-source AI medical imaging framework, to automatically generate patient-specific 3D heart models from CT scans, MRI images, and ultrasound data—compressing a process that once took skilled technicians 4 to 6 hours into just 4 to 6 seconds. The system, built on NVIDIA's MONAI Label and Auto3DSeg tools alongside CHOP's custom SlicerHeart visualization software, is now enabling pediatric cardiologists to rehearse complex surgical procedures and test implanted devices on anatomically precise patient models before entering the operating room.