AI-Enabled Spectral CT Reconstruction

Artificial Intelligence‑enabled spectral CT reconstruction is a new approach to computer‑aided diagnosis. Energy spectral CT uses an additional X-ray at a different energy to obtain more detailed attenuation measurements. This aspect of CT is fundamental information that has not been utilized yet and benefits the diagnosis of the textures or heterogeneities of the lesions. The disclosed technology pairs this new modality with AI-based computer software designed to reconstruct and maximize the image quality.

Technology Overview:

Researchers at Stony Brook University have developed an AI-assisted spectral CT reconstruction coupled with computer-aided diagnosis for identifying lung and colon lesions. This technology comprises two key elements. Firstly, it prioritizes the consideration of structural similarities within the energy spectral channels, encompassing tissue-specific textures and spectral data statistics. This component aims to amplify tissue-specific textures, thereby facilitating the diagnosis of lesion heterogeneities. Secondly, it involves the creation of a texture-based computer-aided diagnosis (CADx) system to effectively harness energy spectral tissue-specific textures for lesion diagnosis. These two components combined allow for improved medical diagnosis using the CT platform.

Advantages:

Maximize potential of CT technology - Provide diagnostic information - Fosters progress in energy spectral CT field

Applications:

Diagnosis of lesion heterogeneities - Provide diagnostic information with the use of CADx

Intellectual Property Summary:

No patent

Stage of Development:

 

Licensing Status:

 

Licensing Potential:

Development partner - Commercial partner - Licensing

Additional Information: