RPI ID: 2018-006-401, 2018-006-601

Innovation Summary:
Researchers at Rensselaer Polytechnic Institute have developed a pair of complementary deep learning technologies that transform how computed tomography (CT) imaging is performed and enhanced. The first system enables adaptive CT imaging by dynamically adjusting scan parameters in real time based on patient-specific data and diagnostic goals. The second invention introduces a post-processing framework that enhances low-dose CT images using a convolutional neural network trained to reduce noise and preserve anatomical detail. Together, these innovations improve diagnostic accuracy, reduce radiation exposure, and integrate seamlessly into existing imaging workflows.

Challenges / Opportunities:
Conventional CT imaging often relies on fixed protocols and post-processing techniques that may not be optimal for every patient or clinical scenario. This can result in unnecessary radiation exposure or compromised image quality. These inventions address those limitations by enabling intelligent, patient-specific imaging strategies and AI-powered image enhancement. They open new opportunities for personalized diagnostics, low-dose imaging, and real-time clinical decision support.

Key Benefits / Advantages:
✔ Adaptive imaging tailored to individual patients and diagnostic needs
✔ Significant reduction in radiation dose without sacrificing image quality
✔ Deep learning-based noise reduction and detail preservation
✔ Real-time optimization and post-acquisition enhancement
✔ Compatible with existing CT systems and clinical workflows

Applications:
• Medical diagnostics and radiology
• Pediatric and geriatric imaging
• Oncology and lung screening
• Emergency and trauma care
• AI-assisted healthcare systems

Keywords:
#Adaptive CT #deep learning #low-dose imaging #medical AI #image enhancement #personalized diagnostics

Intellectual Property:
US issued patents 11,127,175, 11,850,081