Attention-based Generative Latent Replay: A Continual Learning Approach for WSI Analysis

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kumari, Pratibha, Reisenbüchler, Daniel, Bozorgpour, Afshin, Schaadt, Nadine S., Feuerhake, Friedrich, Merhof, Dorit
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913834598400000
author Kumari, Pratibha
Reisenbüchler, Daniel
Bozorgpour, Afshin
Schaadt, Nadine S.
Feuerhake, Friedrich
Merhof, Dorit
author_facet Kumari, Pratibha
Reisenbüchler, Daniel
Bozorgpour, Afshin
Schaadt, Nadine S.
Feuerhake, Friedrich
Merhof, Dorit
contents Whole slide image (WSI) classification has emerged as a powerful tool in computational pathology, but remains constrained by domain shifts, e.g., due to different organs, diseases, or institution-specific variations. To address this challenge, we propose an Attention-based Generative Latent Replay Continual Learning framework (AGLR-CL), in a multiple instance learning (MIL) setup for domain incremental WSI classification. Our method employs Gaussian Mixture Models (GMMs) to synthesize WSI representations and patch count distributions, preserving knowledge of past domains without explicitly storing original data. A novel attention-based filtering step focuses on the most salient patch embeddings, ensuring high-quality synthetic samples. This privacy-aware strategy obviates the need for replay buffers and outperforms other buffer-free counterparts while matching the performance of buffer-based solutions. We validate AGLR-CL on clinically relevant biomarker detection and molecular status prediction across multiple public datasets with diverse centers, organs, and patient cohorts. Experimental results confirm its ability to retain prior knowledge and adapt to new domains, offering an effective, privacy-preserving avenue for domain incremental continual learning in WSI classification.
format Preprint
id arxiv_https___arxiv_org_abs_2505_08524
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Attention-based Generative Latent Replay: A Continual Learning Approach for WSI Analysis
Kumari, Pratibha
Reisenbüchler, Daniel
Bozorgpour, Afshin
Schaadt, Nadine S.
Feuerhake, Friedrich
Merhof, Dorit
Computer Vision and Pattern Recognition
Emerging Technologies
Whole slide image (WSI) classification has emerged as a powerful tool in computational pathology, but remains constrained by domain shifts, e.g., due to different organs, diseases, or institution-specific variations. To address this challenge, we propose an Attention-based Generative Latent Replay Continual Learning framework (AGLR-CL), in a multiple instance learning (MIL) setup for domain incremental WSI classification. Our method employs Gaussian Mixture Models (GMMs) to synthesize WSI representations and patch count distributions, preserving knowledge of past domains without explicitly storing original data. A novel attention-based filtering step focuses on the most salient patch embeddings, ensuring high-quality synthetic samples. This privacy-aware strategy obviates the need for replay buffers and outperforms other buffer-free counterparts while matching the performance of buffer-based solutions. We validate AGLR-CL on clinically relevant biomarker detection and molecular status prediction across multiple public datasets with diverse centers, organs, and patient cohorts. Experimental results confirm its ability to retain prior knowledge and adapt to new domains, offering an effective, privacy-preserving avenue for domain incremental continual learning in WSI classification.
title Attention-based Generative Latent Replay: A Continual Learning Approach for WSI Analysis
topic Computer Vision and Pattern Recognition
Emerging Technologies
url https://arxiv.org/abs/2505.08524