Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Personalized Cell Segmentation: Benchmark and Framework for Reference-Guided Cell Type Segmentation

Bisheng Wang, Jaime S. Cardoso, Lin Wu

Abstract

Accurate cell segmentation is critical for biological and medical imaging studies. Although recent deep learning models have advanced this task, most methods are limited to generic cell segmentation, lacking the ability to differentiate specific cell types. In this work, we introduce the Personalized Cell Segmentation (PerCS) task, which aims to segment all cells of a specific type given a reference cell. To support this task, we establish a benchmark by reorganizing publicly available datasets, yielding 1,372 images and over 110,000 annotated cells. As a pioneering solution, we propose PerCS-DINO, a framework built on the DINOv2 backbone. By integrating image features and reference embeddings via a cross-attention transformer and contrastive learning, PerCS-DINO effectively segments cells matching the reference. Extensive experiments demonstrate the effectiveness of the proposed PerCS-DINO and highlight the challenges of this new task. We expect PerCS to serve as a useful testbed for advancing research in cell-based applications.

Personalized Cell Segmentation: Benchmark and Framework for Reference-Guided Cell Type Segmentation

Abstract

Accurate cell segmentation is critical for biological and medical imaging studies. Although recent deep learning models have advanced this task, most methods are limited to generic cell segmentation, lacking the ability to differentiate specific cell types. In this work, we introduce the Personalized Cell Segmentation (PerCS) task, which aims to segment all cells of a specific type given a reference cell. To support this task, we establish a benchmark by reorganizing publicly available datasets, yielding 1,372 images and over 110,000 annotated cells. As a pioneering solution, we propose PerCS-DINO, a framework built on the DINOv2 backbone. By integrating image features and reference embeddings via a cross-attention transformer and contrastive learning, PerCS-DINO effectively segments cells matching the reference. Extensive experiments demonstrate the effectiveness of the proposed PerCS-DINO and highlight the challenges of this new task. We expect PerCS to serve as a useful testbed for advancing research in cell-based applications.
Paper Structure (15 sections, 2 equations, 3 figures, 1 table)

This paper contains 15 sections, 2 equations, 3 figures, 1 table.

Table of Contents

  1. Introduction
  2. Task Setting and Datasets
  3. Personalized Cell Segmentation (PerCS)
  4. Datasets
  5. PerCS Dataset
  6. PerPanNuke Dataset
  7. Baseline Framework
  8. Adaptation of Existing Works
  9. PerCS-DINO
  10. Experiments
  11. Implementation Details
  12. Evaluation Metrics
  13. Results Comparison
  14. Difficulty Analysis
  15. Conclusion

Figures (3)

  • Figure 1: Framework of generic method, Cellpose-SAM pachitariu2025cellpose. (a) Input images. (b) Image features. (c) Horizontal and vertical vector flows and binary cell mask. (d) Cell mask.
  • Figure 2: Sample images (top row) and corresponding colored annotation masks (bottom row) from the PerCS dataset. Reference targets are outlined in red in the annotation masks.
  • Figure 3: Framework of our proposed method, PerCS-DINO.