ZeroSlide: Is Zero-Shot Classification Adequate for Lifelong Learning in Whole-Slide Image Analysis in the Era of Pathology Vision-Language Foundation Models?

Doanh C. Bui, Pham Hoai Luan, Vu Trung Duong Le, Vu Tuan Hai, Van Duy Tran, Yasuhiko Nakashima

Keywords: lifelong learning, whole slide image analysis, pathology vision-language foundation model

TL;DR: Investigating the performance of zero-shot classification compared to continual learning frameworks for lifelong WSI analysis

Abstract: Lifelong learning for whole-slide images (WSIs) poses the challenge of training a unified model to perform multiple WSI-related tasks, such as cancer subtyping and tumor classification, in a distributed, continual fashion. This is a practical and applicable problem in clinics and hospitals, as WSIs are large, require storage, processing, and transfer time. Training new models whenever new tasks are defined is time-consuming. Recent work has applied regularization‑ and rehearsal‑based methods to this setting. However, the rise of vision‑language foundation models that align diagnostic text with pathology images raises the question: are these models alone sufficient for lifelong WSI learning using zero‑shot classification, or is further investigation into continual‑learning strategies needed to improve performance? The empirical study demonstrates that a well-pretrained pathology vision-language foundation model, when used with a simple zero-shot approach, can achieve competitive performance compared to training-based rehearsal and regularization-based continual learning methods. To our knowledge, this is the first study to compare conventional continual‑learning approaches with vision‑language zero‑shot classification for WSIs. Our source code and experimental results will be available soon.

Submission Number: 1