CellART

CellART is a unified framework for extracting single-cell information from high-resolution ST data. The primary objectives are to accurately delineate boundaries for individual cells and further annotate their cell types. By integrating deep neural networks with probabilistic models, CellART leverages multimodal data, including spatial transcriptomics, staining images, and scRNA-seq references, to perform simultaneous cell segmentation and cell type annotation, thereby optimizing the analytical process.

_images/method.jpg

Note

This project is under active development.

CellART Manuscript

coming soon…

CellART Installation & Usage & Demo

Reproduce CellART Results