Parameter-efficient fine-tuning (PEFT) of Vision Foundation Models for Atypical Mitotic Figure Classification

2025-09-21

AI Summary

Automated classification of atypical mitotic figures (AMFs) in histopathological images remains highly challenging due to subtle morphological variations, severe class imbalance, and substantial inter-annotator disagreement—critical barriers for tumor aggressiveness assessment. To address these issues, we propose a parameter-efficient fine-tuning framework leveraging state-of-the-art vision foundation models (Virchow, Virchow2, and UNI) augmented with Low-Rank Adaptation (LoRA) for lightweight, task-specific adaptation. We systematically investigate the impact of data partitioning strategies and LoRA rank parameters on generalization performance and introduce a three-fold cross-validation ensemble to enhance model robustness. Evaluated on the MIDOG 2025 preliminary test set, our method achieves an 88.37% balanced accuracy, ranking jointly 9th. This work represents the first systematic application of LoRA to fine-grained AMF classification, significantly improving discriminative stability and reproducibility for small-sample, high-noise pathological imagery.

Technology Category

• Machine Learning: Multi-class/Multi-label Learning & Extreme Classification
• Computer Vision: Adversarial Attacks & Robustness
• Humans and AI: Human-in-the-loop Machine Learning

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Atypical mitotic figures (AMFs) are rare abnormal cell divisions associated with tumor aggressiveness and poor prognosis. Their detection remains a significant challenge due to subtle morphological cues, class imbalance, and inter-observer variability among pathologists. The MIDOG 2025 challenge introduced a dedicated track for atypical mitosis classification, enabling systematic evaluation of deep learning methods. In this study, we investigated the use of large vision foundation models, including Virchow, Virchow2, and UNI, with Low-Rank Adaptation (LoRA) for parameter-efficient fine-tuning. We conducted extensive experiments with different LoRA ranks, as well as random and group-based data splits, to analyze robustness under varied conditions. Our best approach, Virchow with LoRA rank 8 and ensemble of three-fold cross-validation, achieved a balanced accuracy of 88.37% on the preliminary test set, ranking joint 9th in the challenge leaderboard. These results highlight the promise of foundation models with efficient adaptation strategies for the classification of atypical mitosis, while underscoring the need for improvements in specificity and domain generalization.

Problem

• Classifying atypical mitotic figures in medical images
• Addressing subtle morphological cues and class imbalance issues
• Improving specificity and domain generalization for cancer prognosis

Innovation

• Parameter-efficient fine-tuning with Low-Rank Adaptation (LoRA)
• Using large vision foundation models like Virchow and UNI
• Employing ensemble of three-fold cross-validation for accuracy