AdaFusion: Prompt-Guided Inference with Adaptive Fusion of Pathology Foundation Models
This addresses the need for more robust and interpretable AI in pathology by mitigating biases in foundation models, though it is incremental as it builds on existing PFMs with a novel fusion method.
The paper tackled the problem of latent biases in pathology foundation models (PFMs) hindering generalisability and transparency by proposing AdaFusion, a prompt-guided inference framework that dynamically integrates multiple PFMs, resulting in consistent performance improvements across three real-world benchmarks for tasks like treatment response prediction and tumour grading.
Pathology foundation models (PFMs) have demonstrated strong representational capabilities through self-supervised pre-training on large-scale, unannotated histopathology image datasets. However, their diverse yet opaque pretraining contexts, shaped by both data-related and structural/training factors, introduce latent biases that hinder generalisability and transparency in downstream applications. In this paper, we propose AdaFusion, a novel prompt-guided inference framework that, to our knowledge, is among the very first to dynamically integrate complementary knowledge from multiple PFMs. Our method compresses and aligns tile-level features from diverse models and employs a lightweight attention mechanism to adaptively fuse them based on tissue phenotype context. We evaluate AdaFusion on three real-world benchmarks spanning treatment response prediction, tumour grading, and spatial gene expression inference. Our approach consistently surpasses individual PFMs across both classification and regression tasks, while offering interpretable insights into each model's biosemantic specialisation. These results highlight AdaFusion's ability to bridge heterogeneous PFMs, achieving both enhanced performance and interpretability of model-specific inductive biases.