IRS: Incremental Relationship-guided Segmentation for Digital Pathology

2025-05-28

AI Summary

Continuous segmentation of digital pathology whole-slide images faces challenges from sequential acquisition, partial annotations, and distributional shifts—particularly out-of-distribution (OOD) novel lesions (e.g., previously unseen diseases or cellular phenotypes). Method: We propose a spatio-temporal OOD continual learning paradigm. Our approach introduces an incremental universal proposition matrix to model cross-class anatomical relationships, integrated with multi-scale feature alignment and a panoptic segmentation framework, enabling zero-shot class expansion and joint segmentation at region, unit, and cellular levels. Results: Evaluated on kidney pathology data, our method significantly improves generalization across institutions, imaging devices, and disease types. Under partial annotation settings, it achieves accurate OOD lesion identification and multi-scale structural segmentation. This advances the development of clinically deployable, adaptive AI systems for digital pathology.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Life-Long and Continual Learning
• Search and Optimization: Learning to Search

Application Category

• Security and Privacy: Data transparency and provenance
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Continual learning is rapidly emerging as a key focus in computer vision, aiming to develop AI systems capable of continuous improvement, thereby enhancing their value and practicality in diverse real-world applications. In healthcare, continual learning holds great promise for continuously acquired digital pathology data, which is collected in hospitals on a daily basis. However, panoramic segmentation on digital whole slide images (WSIs) presents significant challenges, as it is often infeasible to obtain comprehensive annotations for all potential objects, spanning from coarse structures (e.g., regions and unit objects) to fine structures (e.g., cells). This results in temporally and partially annotated data, posing a major challenge in developing a holistic segmentation framework. Moreover, an ideal segmentation model should incorporate new phenotypes, unseen diseases, and diverse populations, making this task even more complex. In this paper, we introduce a novel and unified Incremental Relationship-guided Segmentation (IRS) learning scheme to address temporally acquired, partially annotated data while maintaining out-of-distribution (OOD) continual learning capacity in digital pathology. The key innovation of IRS lies in its ability to realize a new spatial-temporal OOD continual learning paradigm by mathematically modeling anatomical relationships between existing and newly introduced classes through a simple incremental universal proposition matrix. Experimental results demonstrate that the IRS method effectively handles the multi-scale nature of pathological segmentation, enabling precise kidney segmentation across various structures (regions, units, and cells) as well as OOD disease lesions at multiple magnifications. This capability significantly enhances domain generalization, making IRS a robust approach for real-world digital pathology applications.

Problem

• Addresses continual learning for digital pathology segmentation
• Handles partially annotated multi-scale pathological data
• Enhances domain generalization for diverse disease lesions

Innovation

• Incremental Relationship-guided Segmentation (IRS) scheme
• Spatial-temporal OOD continual learning paradigm
• Incremental universal proposition matrix modeling