One Click per Cell Type Suffices: Training-free Group Interaction for Cell Instance Segmentation

2026-05-28

AI Summary

This work addresses the significant performance degradation of conventional cell instance segmentation models on out-of-distribution cell types and the inefficiency of existing interactive methods that require per-instance prompting. The authors propose a novel Group Prompting paradigm, enabling interaction at the cell-type level rather than the instance level: a single click per cell type suffices to segment all instances of that type. Built upon a Chain-of-Prompts framework, the method recursively expands reliable prompt points on the frozen multi-scale image features of SAM, leveraging their inherent ability to aggregate similar cells. This reduces interaction complexity from O(N) to O(T), where N is the number of instances and T the number of types. Experiments demonstrate that with just one click per type, the approach retains over 90% of per-instance performance on three annotated datasets and exceeds 99% on four homogeneous datasets—surpassing fully supervised baselines without any additional training.

Abstract

Cell instance segmentation models trained on cell-specific datasets suffer severe performance drops on out-of-distribution cell types, while interactive foundation models overcome this through per-instance prompting at a cost that is prohibitively expensive for histopathology images containing hundreds to thousands of densely packed instances. We introduce Group Prompting, a new paradigm that shifts interactive segmentation from per-instance $O(N)$ to per-type $O(T)$, where a single click per cell type suffices to segment all instances of that type. Our key observation is that the frozen image encoder of the Segment Anything Model (SAM) already clusters same-type cells in its feature space before any prompt is given. Exploiting this property, we propose Chain-of-Prompts (CoP), a training-free framework that recursively expands a single user click by (1) identifying reliable same-type locations through non-parametric gating of multi-scale encoder features, and (2) selecting the most spatially distant reliable point as the next prompt to maximize coverage. On three cell-type-annotated benchmarks, CoP with one click per type retains over 90% of per-instance performance and surpasses fully-supervised methods without any additional training. On four morphologically homogeneous benchmarks, a single click retains over 99%. Project Page: https://shjo-april.github.io/Chain-of-Prompts/

Problem

• cell instance segmentation
• out-of-distribution generalization
• interactive segmentation
• histopathology images
• prompting

Innovation

• Group Prompting
• Chain-of-Prompts
• Training-free
• Cell Instance Segmentation
• Interactive Segmentation