RPI ID: 2018-020-401 / 2018-020-601

Innovation Summary:
This invention introduces a neural network-based corrector for photon-counting computed tomography (PCCT) systems. Designed to address spectral distortions and detector non-linearities, the system processes raw photon data to generate corrected images with enhanced spectral fidelity and quantitative accuracy. Optimized for real-time performance, the model is trained on both synthetic and clinical datasets and is compatible with existing CT workflows. This AI-driven approach supports the clinical translation of photon-counting CT by improving image quality and diagnostic reliability.

Challenges / Opportunities:
Photon-counting CT offers superior resolution and spectral imaging capabilities but is hindered by detector-related distortions and spectral inaccuracies. Traditional correction methods are often rigid and limited in adaptability. This invention provides a scalable, intelligent solution that adapts to diverse imaging conditions, enabling broader clinical adoption of PCCT technology and supporting precision diagnostics in oncology, cardiology, and beyond.

Key Benefits / Advantages:
✔ Spectral correction for improved material differentiation and contrast
✔ Real-time processing suitable for clinical environments
✔ AI-driven adaptability across imaging conditions
✔ Seamless integration with commercial CT platforms
✔ Supports quantitative imaging and diagnostic consistency

Applications:
• Oncology and cardiovascular diagnostics
• Spectral CT imaging
• Advanced radiology systems
• Medical imaging research and development

Keywords:
#photoncountingCT #spectralcorrection #neuralnetworks #AIimaging #CTenhancement #quantitativeimaging

Intellectual Property:
US Issued Patents 11,448,778 B2 and 11,686,863 B2