A foundation model for generalizable disease diagnosis in chest X-ray images

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Xu, Lijian, Ni, Ziyu, Sun, Hao, Li, Hongsheng, Zhang, Shaoting
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866914970156924928
author Xu, Lijian
Ni, Ziyu
Sun, Hao
Li, Hongsheng
Zhang, Shaoting
author_facet Xu, Lijian
Ni, Ziyu
Sun, Hao
Li, Hongsheng
Zhang, Shaoting
contents Medical artificial intelligence (AI) is revolutionizing the interpretation of chest X-ray (CXR) images by providing robust tools for disease diagnosis. However, the effectiveness of these AI models is often limited by their reliance on large amounts of task-specific labeled data and their inability to generalize across diverse clinical settings. To address these challenges, we introduce CXRBase, a foundational model designed to learn versatile representations from unlabelled CXR images, facilitating efficient adaptation to various clinical tasks. CXRBase is initially trained on a substantial dataset of 1.04 million unlabelled CXR images using self-supervised learning methods. This approach allows the model to discern meaningful patterns without the need for explicit labels. After this initial phase, CXRBase is fine-tuned with labeled data to enhance its performance in disease detection, enabling accurate classification of chest diseases. CXRBase provides a generalizable solution to improve model performance and alleviate the annotation workload of experts to enable broad clinical AI applications from chest imaging.
format Preprint
id arxiv_https___arxiv_org_abs_2410_08861
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle A foundation model for generalizable disease diagnosis in chest X-ray images
Xu, Lijian
Ni, Ziyu
Sun, Hao
Li, Hongsheng
Zhang, Shaoting
Image and Video Processing
Computer Vision and Pattern Recognition
Medical artificial intelligence (AI) is revolutionizing the interpretation of chest X-ray (CXR) images by providing robust tools for disease diagnosis. However, the effectiveness of these AI models is often limited by their reliance on large amounts of task-specific labeled data and their inability to generalize across diverse clinical settings. To address these challenges, we introduce CXRBase, a foundational model designed to learn versatile representations from unlabelled CXR images, facilitating efficient adaptation to various clinical tasks. CXRBase is initially trained on a substantial dataset of 1.04 million unlabelled CXR images using self-supervised learning methods. This approach allows the model to discern meaningful patterns without the need for explicit labels. After this initial phase, CXRBase is fine-tuned with labeled data to enhance its performance in disease detection, enabling accurate classification of chest diseases. CXRBase provides a generalizable solution to improve model performance and alleviate the annotation workload of experts to enable broad clinical AI applications from chest imaging.
title A foundation model for generalizable disease diagnosis in chest X-ray images
topic Image and Video Processing
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2410.08861