3D Foundation AI Model for Generalizable Disease Detection in Head Computed Tomography

2025-02-04

AI Summary

This paper addresses the challenges of poor generalization and scarcity of high-quality annotations in multi-disease detection from cranial CT images. To this end, we propose FM-CT—the first 3D self-supervised foundation model specifically designed for cranial CT. Leveraging 362,000 unlabeled, non-contrast-enhanced 3D CT volumes, FM-CT introduces a novel 3D structured self-supervision paradigm that jointly integrates self-distillation-based discriminative learning and masked voxel modeling. The resulting model achieves significantly improved in-distribution (ID) and out-of-distribution (OOD) generalization, particularly excelling in few-shot downstream tasks. Evaluated on an internal dataset and three multi-center external cohorts, FM-CT consistently outperforms both scratch-trained models and existing 3D CT foundation models across diverse diagnostic tasks. It thereby establishes a new state-of-the-art benchmark for 3D AI-driven analysis of cranial CT.

Technology Category

• Computer Vision: 3D Computer Vision
• Machine Learning: Deep Neural Architectures and Foundation Models
• Constraint Satisfaction and Optimization: Other Foundations of Constraint Satisfaction

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Head computed tomography (CT) imaging is a widely-used imaging modality with multitudes of medical indications, particularly in assessing pathology of the brain, skull, and cerebrovascular system. It is commonly the first-line imaging in neurologic emergencies given its rapidity of image acquisition, safety, cost, and ubiquity. Deep learning models may facilitate detection of a wide range of diseases. However, the scarcity of high-quality labels and annotations, particularly among less common conditions, significantly hinders the development of powerful models. To address this challenge, we introduce FM-CT: a Foundation Model for Head CT for generalizable disease detection, trained using self-supervised learning. Our approach pre-trains a deep learning model on a large, diverse dataset of 361,663 non-contrast 3D head CT scans without the need for manual annotations, enabling the model to learn robust, generalizable features. To investigate the potential of self-supervised learning in head CT, we employed both discrimination with self-distillation and masked image modeling, and we construct our model in 3D rather than at the slice level (2D) to exploit the structure of head CT scans more comprehensively and efficiently. The model's downstream classification performance is evaluated using internal and three external datasets, encompassing both in-distribution (ID) and out-of-distribution (OOD) data. Our results demonstrate that the self-supervised foundation model significantly improves performance on downstream diagnostic tasks compared to models trained from scratch and previous 3D CT foundation models on scarce annotated datasets. This work highlights the effectiveness of self-supervised learning in medical imaging and sets a new benchmark for head CT image analysis in 3D, enabling broader use of artificial intelligence for head CT-based diagnosis.

Problem

• Develops 3D AI for head CT disease detection
• Addresses scarcity of high-quality medical annotations
• Uses self-supervised learning for robust feature extraction

Innovation

• 3D self-supervised learning
• Foundation Model for Head CT
• Masked image modeling