A Non-contrast Head CT Foundation Model for Comprehensive Neuro-Trauma Triage

2025-02-28

AI Summary

To address the global shortage of radiologists and the surging demand for rapid interpretation of non-contrast head CT scans in emergency settings, this study develops the first 3D foundation model specifically designed for non-contrast head CT. The model enables fully automated, high-accuracy triage across 16 neurotrauma categories—including intracranial hemorrhage subtypes, midline shift, and cerebral edema. Methodologically, we propose: (1) an LLM-driven automatic multi-label annotation framework; (2) a dual-path pretraining paradigm integrating neuroanatomical priors with lesion segmentation; and (3) a neuro-specific multimodal fine-tuning strategy. Built upon a 3D CNN architecture, the model jointly learns anatomical–pathological representations. On an independent test set, it achieves a mean AUC of 0.861—significantly outperforming CT-CLIP (p < 0.01) and meeting clinically actionable triage accuracy.

Technology Category

• Machine Learning: Deep Neural Architectures and Foundation Models
• Computer Vision: Large Vision Models
• Cognitive Modeling & Cognitive Systems: Other Foundations of Cognitive Modeling & Systems

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Recent advancements in AI and medical imaging offer transformative potential in emergency head CT interpretation for reducing assessment times and improving accuracy in the face of an increasing request of such scans and a global shortage in radiologists. This study introduces a 3D foundation model for detecting diverse neuro-trauma findings with high accuracy and efficiency. Using large language models (LLMs) for automatic labeling, we generated comprehensive multi-label annotations for critical conditions. Our approach involved pretraining neural networks for hemorrhage subtype segmentation and brain anatomy parcellation, which were integrated into a pretrained comprehensive neuro-trauma detection network through multimodal fine-tuning. Performance evaluation against expert annotations and comparison with CT-CLIP demonstrated strong triage accuracy across major neuro-trauma findings, such as hemorrhage and midline shift, as well as less frequent critical conditions such as cerebral edema and arterial hyperdensity. The integration of neuro-specific features significantly enhanced diagnostic capabilities, achieving an average AUC of 0.861 for 16 neuro-trauma conditions. This work advances foundation models in medical imaging, serving as a benchmark for future AI-assisted neuro-trauma diagnostics in emergency radiology.

Problem

• Develops a 3D AI model for neuro-trauma detection
• Improves accuracy and speed in emergency head CT scans
• Addresses radiologist shortage with advanced diagnostic tools

Innovation

• 3D foundation model for neuro-trauma detection
• Large language models for automatic labeling
• Multimodal fine-tuning enhances diagnostic accuracy