Maskless 2D/3D Artificial Subtraction Angiography

LUNU2019-160

INVENTORS

• Sameer Ansari*
• Donald Cantrell*

SHORT DESCRIPTION

This invention leverages machine learning to enhance angiographic imaging by removing bone and soft tissue interference without a pre-injection digital mask, providing improved visualization of vasculature in real-time and accommodating patient motion.

BACKGROUND

Conventional digital subtraction angiography techniques depend on acquiring a mask image prior to contrast administration, making them highly susceptible to motion artifacts when there are even minor patient movements between image acquisitions. This reliance on matched pre- and post-contrast images not only heightens the risk of registration errors but also increases radiation exposure and limits image quality, particularly in dynamic or complex anatomical regions. Additionally, current approaches struggle to isolate vascular features from surrounding bone and soft tissues in real-time, posing significant challenges for accurate, high-quality visualization in both two-dimensional and three-dimensional imaging scenarios.

ABSTRACT

This invention introduces a deep learning-based system for maskless artificial subtraction angiography, improving the clarity and safety of blood vessel imaging. Unlike traditional methods that require a pre-contrast mask image and are prone to motion-related artifacts, this novel technology uses advanced algorithms to directly isolate blood vessels from raw imaging data. It processes 2D and 3D angiographic images in real-time, even accounting for patient movement, to deliver high-quality, motion-stabilized images without additional radiation exposure. Compatible with various imaging devices like CT, MRI, and rotational angiography, this system offers a faster, safer, and more versatile solution for visualizing vascular structures in different parts of the body.

APPLICATIONS

• AI-Enhanced Diagnostic Imaging Software Integration
  • Deep learning-based, maskless subtraction process improves vasculature image clarity across multiple imaging modalities.

ADVANTAGES

• Image Quality Enhancement
  • Significantly reducing motion artifacts that commonly degrade image quality in traditional digital subtraction angiography (DSA).
• Real-time Processing Capability
  • With processing speeds of at least 16 frames per second, this system enables immediate image generation.
• Multimodality Versatility
  • Compatibility with various imaging modalities—including 3D rotational angiography, CT, and MRI
• Cost and Radiation Exposure Reduction
  • Higher quality scans reduces frequence and cost of repeat scans and limits radiation exposure.

IP STATUS

Pending US Patent