CardioNVT: a deep learning framework for cardiomyocyte nuclear volume tracking and ploidy Analysis. (IMAGE)
Caption
CardioNVT facilitates cardiomyocyte nuclear ploidy inference from DAPI-only stained images of cardiac tissue sections. By combining UNet++ and SAM2, the framework enables single-plane cardiomyocyte nuclear segmentation and cross-plane nuclear tracking, allowing nuclear volume estimation from serial tissue sections. The resulting nuclear volume distributions support ploidy analysis and thus provide a new solution for investigating ploidy alterations during cardiac disease-associated remodeling.
Credit
©Science Bulletin
Usage Restrictions
use with credit.
License
Original content