RedDino: A foundation model for red blood cell analysis

2025-08-11

AI Summary

Current red blood cell (RBC) image analysis lacks a general-purpose, robust AI foundation model, hindering precise diagnosis of hematologic disorders. To address this, we propose the first self-supervised foundation model specifically designed for RBC morphological analysis. Our method innovatively customizes the DINOv2 framework to accommodate RBC-specific characteristics—including scale variability, low contrast, and high morphological deformability. Trained on 1.25 million multi-modal RBC images (including microscope, digital pathology, and peripheral blood smear acquisitions) without any human annotations, the model achieves state-of-the-art performance on fine-grained RBC shape classification via linear probing and nearest-neighbor evaluation. It further demonstrates exceptional cross-device and cross-institution generalization. This work establishes a transferable, scalable visual foundation model to support clinical decision-making in hematologic disease diagnosis.

Technology Category

• Computer Vision: Other Foundations of Computer Vision
• Machine Learning: Deep Neural Architectures and Foundation Models
• Humans and AI: Other Foundations of Human Computation & AI

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Red blood cells (RBCs) are essential to human health, and their precise morphological analysis is important for diagnosing hematological disorders. Despite the promise of foundation models in medical diagnostics, comprehensive AI solutions for RBC analysis remain scarce. We present RedDino, a self-supervised foundation model designed for RBC image analysis. RedDino uses an RBC-specific adaptation of the DINOv2 self-supervised learning framework and is trained on a curated dataset of 1.25 million RBC images from diverse acquisition modalities and sources. Extensive evaluations show that RedDino outperforms existing state-of-the-art models on RBC shape classification. Through assessments including linear probing and nearest neighbor classification, we confirm its strong feature representations and generalization ability. Our main contributions are: (1) a foundation model tailored for RBC analysis, (2) ablation studies exploring DINOv2 configurations for RBC modeling, and (3) a detailed evaluation of generalization performance. RedDino addresses key challenges in computational hematology by capturing nuanced morphological features, advancing the development of reliable diagnostic tools. The source code and pretrained models for RedDino are available at https://github.com/Snarci/RedDino, and the pretrained models can be downloaded from our Hugging Face collection at https://huggingface.co/collections/Snarcy/reddino-689a13e29241d2e5690202fc

Problem

• Develops a foundation model for precise RBC morphological analysis
• Addresses scarcity of comprehensive AI solutions for RBC diagnostics
• Enhances feature representation for reliable hematological disorder detection

Innovation

• Self-supervised foundation model for RBC analysis
• Adapts DINOv2 framework for RBC-specific learning
• Trained on 1.25 million diverse RBC images