Adapting Self-Supervised Learning for Computational Pathology
This work addresses the challenge of limited labeled data in computational pathology, offering a tailored approach that enhances SSL for medical image analysis, though it is incremental as it builds on existing SSL methods.
The paper tackled the problem of adapting self-supervised learning (SSL) for computational pathology by modifying the DINOv2 algorithm with domain-specific augmentations, regularization, and position encodings, resulting in improved performance on several benchmarks.
Self-supervised learning (SSL) has emerged as a key technique for training networks that can generalize well to diverse tasks without task-specific supervision. This property makes SSL desirable for computational pathology, the study of digitized images of tissues, as there are many target applications and often limited labeled training samples. However, SSL algorithms and models have been primarily developed in the field of natural images and whether their performance can be improved by adaptation to particular domains remains an open question. In this work, we present an investigation of modifications to SSL for pathology data, specifically focusing on the DINOv2 algorithm. We propose alternative augmentations, regularization functions, and position encodings motivated by the characteristics of pathology images. We evaluate the impact of these changes on several benchmarks to demonstrate the value of tailored approaches.