Artificial Intelligence-Assisted Low-Dose High Atomic Number Contrast Agent for Ultrahigh-Resolution Computed...

Area of Science:

• Nanotechnology
• Biomedical Imaging
• Artificial Intelligence

Background:

• Minimizing contrast agent dosage in computed tomography (CT) while maintaining high resolution is a significant challenge.
• Developing safer and more effective contrast agents is crucial for advanced medical imaging.

Purpose of the Study:

• To develop an artificial intelligence-assisted, low-dose, high atomic number contrast agent for ultrahigh-resolution CT imaging.
• To demonstrate the efficacy of PEGylated hafnium oxide nanoparticles (DA-HfO2 NPs) for enhanced CT angiography.

Main Methods:

• Synthesis of high-quality PEGylated hafnium oxide nanoparticles (DA-HfO2 NPs) with specific properties (36% hafnium content, 13.5 nm size, 161.9 min half-life).
• Evaluation of DA-HfO2 NPs for high-resolution vascular imaging at both high and low doses (25% of standard).
• Integration of artificial intelligence algorithms to enhance imaging quality at reduced contrast agent concentrations.

Main Results:

• High-dose DA-HfO2 NPs enabled ultrahigh-resolution vascular imaging (0.15 mm spatial resolution) for at least 60 minutes.
• AI-assisted low-dose DA-HfO2 NPs achieved imaging quality comparable to high-dose agents in contrast density and spatial resolution.
• The low-dose approach significantly enhanced biosafety while maintaining diagnostic image quality.

Conclusions:

• The proposed strategy offers a promising method for high-resolution CT imaging with reduced contrast agent doses.
• AI-integrated low-dose hafnium oxide nanoparticles represent a sensitive and safe advancement for CT angiography.
• This approach addresses the critical need for improved contrast-enhanced CT while prioritizing patient safety.