Self-Supervised Learning for Anatomy in Chest Radiographs

This paper addresses a critical gap in medical imaging by leveraging self-supervised learning to build foundation models that understand human anatomy. The core idea is to exploit the inherent structure and consistency of anatomical features within chest radiographs, leading to more robust and transferable representations compared to existing methods. The focus on multiple perspectives and the use of anatomical principles as a supervision signal are key innovations.

• •Proposes Lamps, a self-supervised learning approach for chest radiograph analysis.
• •Utilizes anatomical consistency, coherence, and hierarchy as supervision signals.
• •Demonstrates superior performance in fine-tuning and emergent property analysis compared to baselines.
• •Aims to create foundation models aligned with the structure of human anatomy.