论文阅读记录

Bering

• 标题：Bering: joint cell segmentation and annotation for spatial transcriptomics with transferred graph embeddings
• 作者：Harvard University
• 发表会议/期刊：Nature Communications
• 年份：2025
• 链接：Bering

主要内容简介

• Motivation:
  Some tissues have densely packed cells with unclear boundaries, making it difficult to perform accurate segmentation

• Task:
  Cell segmentation and annotation for spatial transcriptomics

方法与创新点

• 总体框架：

• 方法概述：

  1. 图构建，NGC
     

  2. 图卷积和全连接网络
     

  3. 节点分类
     

  4. 边嵌入由三部分组成
     - node representation
     
     - distance kernels($\mu_d, \sigma_d$可学习参数)
     
     - image representation(使用CNN+SPP,边作为对角线图像)
     
     Edge embedding:
     
     Edge classification（预测边标签，是否同一个细胞内）:
     
     聚类算法划分细胞（Leiden算法）:
     

  5. Other details

     • cell type annotation details
     • transfer learning
     • NGC Matrix
     • Generation of validation data: FOVs