Lamps: Learning Anatomy from Multiple Perspectives via Self-supervision in Chest Radiographs

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhou, Ziyu, Luo, Haozhe, Taher, Mohammad Reza Hosseinzadeh, Pang, Jiaxuan, Ding, Xiaowei, Gotway, Michael B., Liang, Jianming
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909979636662272
author Zhou, Ziyu
Luo, Haozhe
Taher, Mohammad Reza Hosseinzadeh
Pang, Jiaxuan
Ding, Xiaowei
Gotway, Michael B.
Liang, Jianming
author_facet Zhou, Ziyu
Luo, Haozhe
Taher, Mohammad Reza Hosseinzadeh
Pang, Jiaxuan
Ding, Xiaowei
Gotway, Michael B.
Liang, Jianming
contents Foundation models have been successful in natural language processing and computer vision because they are capable of capturing the underlying structures (foundation) of natural languages. However, in medical imaging, the key foundation lies in human anatomy, as these images directly represent the internal structures of the body, reflecting the consistency, coherence, and hierarchy of human anatomy. Yet, existing self-supervised learning (SSL) methods often overlook these perspectives, limiting their ability to effectively learn anatomical features. To overcome the limitation, we built Lamps (learning anatomy from multiple perspectives via self-supervision) pre-trained on large-scale chest radiographs by harmoniously utilizing the consistency, coherence, and hierarchy of human anatomy as the supervision signal. Extensive experiments across 10 datasets evaluated through fine-tuning and emergent property analysis demonstrate Lamps' superior robustness, transferability, and clinical potential when compared to 10 baseline models. By learning from multiple perspectives, Lamps presents a unique opportunity for foundation models to develop meaningful, robust representations that are aligned with the structure of human anatomy.
format Preprint
id arxiv_https___arxiv_org_abs_2512_22872
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Lamps: Learning Anatomy from Multiple Perspectives via Self-supervision in Chest Radiographs
Zhou, Ziyu
Luo, Haozhe
Taher, Mohammad Reza Hosseinzadeh
Pang, Jiaxuan
Ding, Xiaowei
Gotway, Michael B.
Liang, Jianming
Computer Vision and Pattern Recognition
Foundation models have been successful in natural language processing and computer vision because they are capable of capturing the underlying structures (foundation) of natural languages. However, in medical imaging, the key foundation lies in human anatomy, as these images directly represent the internal structures of the body, reflecting the consistency, coherence, and hierarchy of human anatomy. Yet, existing self-supervised learning (SSL) methods often overlook these perspectives, limiting their ability to effectively learn anatomical features. To overcome the limitation, we built Lamps (learning anatomy from multiple perspectives via self-supervision) pre-trained on large-scale chest radiographs by harmoniously utilizing the consistency, coherence, and hierarchy of human anatomy as the supervision signal. Extensive experiments across 10 datasets evaluated through fine-tuning and emergent property analysis demonstrate Lamps' superior robustness, transferability, and clinical potential when compared to 10 baseline models. By learning from multiple perspectives, Lamps presents a unique opportunity for foundation models to develop meaningful, robust representations that are aligned with the structure of human anatomy.
title Lamps: Learning Anatomy from Multiple Perspectives via Self-supervision in Chest Radiographs
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2512.22872