DA-SSL: self-supervised domain adaptor to leverage foundational models in turbt histopathology slides

2025-12-15

AI Summary

To address domain shift in Pathology Foundation Models (PFMs) caused by tissue fragmentation and electrocautery artifacts in transurethral resection of bladder tumor (TURBT) specimens, this work proposes DA-SSL—a self-supervised domain adaptation adapter that requires no PFM fine-tuning. DA-SSL integrates multi-instance learning (MIL) with contrastive self-supervised learning to achieve lightweight, non-invasive feature-space alignment, preserving pre-trained knowledge while accommodating the scarcity of clinical TURBT samples. It enables robust multi-center generalization: under five-fold cross-validation, it achieves an AUC of 0.77±0.04; on external validation, it attains accuracy=0.84, sensitivity=0.71, and specificity=0.91—significantly improving prediction of muscle-invasive bladder cancer (MIBC) treatment response. To our knowledge, DA-SSL is the first self-supervised domain adaptation framework tailored to TURBT histopathology that operates without modifying PFM parameters.

Technology Category

• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Computer Vision: Multi-modal Vision
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization

Abstract

Recent deep learning frameworks in histopathology, particularly multiple instance learning (MIL) combined with pathology foundational models (PFMs), have shown strong performance. However, PFMs exhibit limitations on certain cancer or specimen types due to domain shifts - these cancer types were rarely used for pretraining or specimens contain tissue-based artifacts rarely seen within the pretraining population. Such is the case for transurethral resection of bladder tumor (TURBT), which are essential for diagnosing muscle-invasive bladder cancer (MIBC), but contain fragmented tissue chips and electrocautery artifacts and were not widely used in publicly available PFMs. To address this, we propose a simple yet effective domain-adaptive self-supervised adaptor (DA-SSL) that realigns pretrained PFM features to the TURBT domain without fine-tuning the foundational model itself. We pilot this framework for predicting treatment response in TURBT, where histomorphological features are currently underutilized and identifying patients who will benefit from neoadjuvant chemotherapy (NAC) is challenging. In our multi-center study, DA-SSL achieved an AUC of 0.77+/-0.04 in five-fold cross-validation and an external test accuracy of 0.84, sensitivity of 0.71, and specificity of 0.91 using majority voting. Our results demonstrate that lightweight domain adaptation with self-supervision can effectively enhance PFM-based MIL pipelines for clinically challenging histopathology tasks. Code is Available at https://github.com/zhanghaoyue/DA_SSL_TURBT.

Problem

• Addresses domain shifts in pathology foundational models for rare cancer types.
• Enhances PFM-based MIL pipelines for challenging histopathology tasks via self-supervised adaptation.
• Improves prediction of neoadjuvant chemotherapy response in bladder cancer using TURBT slides.

Innovation

• Self-supervised domain adaptation realigns pretrained features
• Lightweight adaptor works without fine-tuning the foundational model
• Enhances multiple instance learning for challenging histopathology tasks