AdvDINO: Domain-Adversarial Self-Supervised Representation Learning for Spatial Proteomics
This work addresses domain shift issues in biomedical imaging, such as batch effects in spatial proteomics, which is an incremental advancement with potential broader applications in other imaging domains.
The paper tackled the problem of domain shift in self-supervised learning for biomedical imaging by introducing AdvDINO, a domain-adversarial framework that mitigates slide-specific biases in multiplex immunofluorescence images, resulting in improved survival prediction and the uncovering of phenotype clusters with prognostic significance across over 5.46 million image tiles.
Self-supervised learning (SSL) has emerged as a powerful approach for learning visual representations without manual annotations. However, the robustness of standard SSL methods to domain shift -- systematic differences across data sources -- remains uncertain, posing an especially critical challenge in biomedical imaging where batch effects can obscure true biological signals. We present AdvDINO, a domain-adversarial self-supervised learning framework that integrates a gradient reversal layer into the DINOv2 architecture to promote domain-invariant feature learning. Applied to a real-world cohort of six-channel multiplex immunofluorescence (mIF) whole slide images from non-small cell lung cancer patients, AdvDINO mitigates slide-specific biases to learn more robust and biologically meaningful representations than non-adversarial baselines. Across $>5.46$ million mIF image tiles, the model uncovers phenotype clusters with distinct proteomic profiles and prognostic significance, and improves survival prediction in attention-based multiple instance learning. While demonstrated on mIF data, AdvDINO is broadly applicable to other imaging domains -- including radiology, remote sensing, and autonomous driving -- where domain shift and limited annotated data hinder model generalization and interpretability.