费城儿童医院用NVIDIA开源AI把心脏扫描转成3D模型，耗时从4小时缩至秒级
Children’s Hospital of Philadelphia is using NVIDIA-backed open source AI to turn medical scans into detailed models of hearts. What once took a skilled researcher four hours can now happen in seconds. These models help doctors understand complex heart defects and plan repairs around each child’s anatomy. Researchers are also developing physics simulations to predict how medical devices would fit and interact with heart tissue before a procedure. That real-time workflow is still being developed, but anatomical modeling already supports care at the hospital. Open source software lets hospitals share improvements and build better tools for children with unique hearts. The next step could be testing devices inside a virtual heart before placing them in a real one.
费城儿童医院使用NVIDIA支持的开源AI框架MONAI，把儿童心脏医学扫描转成详细的三维解剖模型，原本需要熟练研究员4小时的工作现在几秒内完成。团队用既往扫描与专家制作的模型配对训练神经网络，让AI完成解剖结构的分割识别，帮助医生理解复杂心脏缺陷并按每个孩子的解剖结构规划修复方案。
视频展示了心脏3D模型和分割流程，适合做案例解读，重点讲4小时工作缩短到秒级以及开源框架MONAI的复用方式。
粉丝 8万
The latest in AI and robotics, explained Founder & CEO of The Rundown AI Zurich 🇨🇭 / NYC 🇺🇸
- 浏览
- 726
- 点赞
- 37
- 评论
- 1
- 分享
- 4
- 收藏
- 7
数字来自平台，为最近一次抓取时的数据。
以下是视频口播的自动转写，可能有错字。
This 3d model helps surgeons locate a hole in a child's heart, and now the children's hospital of philadelphia is using open source ai tools from nvidia to produce heart models in seconds.
The framework is called monai, and it helps turn medical scans into detailed maps of a child's heart anatomy. The team trains neural networks using previous scans paired with models created by specialists.
The ai learns to identify and separate anatomical structures, which is a process called segmentation. And get this, that work previously took a skilled researcher 4 hours, and it can now happen in seconds.
That matters because children with congenital heart defects can have dramatically different anatomy, while the devices available to treat them come in standard designs.
The team is also developing physics simulations using nvidia warp and newton to predict how the device would expand inside a particular heart and press against surrounding tissue.
The ultimate goal here is to let doctors compare the fit. Before the procedure, and now that real time simulation workflow is still being developed, but the modeling already helps the hospital plan complex operations.
And here's the coolest part, because that software is open source, hospitals can share. improvements building tools together for children whose hearts require an individual approach.
If you want to stay in the loop with the latest in AI and robotics, follow for more.
来源：TikTok · Rowan Cheung · tiktok.com