Spectral CT Material Decomposition from Standard Single-Energy X-Ray Scans

Summary

This technology enables multi-material image reconstruction from conventional single-spectrum X-ray tomography data, including standard CT and tomosynthesis, without requiring dual-energy CT or photon-counting CT hardware. The approach can generate basis-material images and virtual monochromatic images from data acquired using standard X-ray sources and detectors.

Unmet need: Advanced material-specific CT imaging without specialized spectral CT hardware.

Multi-material imaging is valuable in clinical CT because it can differentiate and quantify materials such as water, adipose tissue, glandular tissue, tumor tissue, iodine contrast agents, calcium, bone, and metals. Today, these capabilities typically require expensive specialized equipment such as dual-energy CT, multi-spectral CT, or photon-counting CT. When compared to standard CT these instruments are significantly more complex leading to a greater burden on operators and increase maintenance cost and requirements.

Proposed solution: Software-based multi-basis reconstruction from conventional single-spectrum X-ray tomography data

This technology is an image reconstruction method that directly inverts the nonlinear data model of standard polychromatic CT to recover multiple basis-material images from conventional single-spectrum scan data. The dynamic non-convex primal-dual algorithm uses basis-region modeling to reduce the number of unknowns and a volume-conservation constraint to augment the available information, enabling reconstruction of multi-basis images and virtual monochromatic images from standard CT data. In studies performed at the University of Chicago, the method reconstructed basis images such as air, water, bone, iodine, calcium, and adipose from simulated and real conventional CT datasets, with results visually comparable to dual-energy CT reconstructions in phantom studies.

Key Advantages

• No specialized hardware required
• Potential alternative to DECT and PCCT
• Compatible with existing standard CT and tomosynthesis workflows
• Generates multi-basis-material images
• Can produce virtual monochromatic images
• May reduce imaging cost, scan complexity, and hardware dependence
• Supports quantitative material assessment
• Extendable to cone-beam CT, sparse-view CT, and limited-angular-range CT applications

Potential Applications

• Standard CT material decomposition
• CT beam-hardening correction
• Virtual monochromatic imaging
• Contrast-agent quantification, including iodine imaging
• Calcium and bone characterization
• Oncology imaging and tumor/tissue differentiation
• Breast tomosynthesis material imaging
• Cone-beam CT and interventional imaging
• Retrospective enhancement of conventional CT protocols
• Cost-effective spectral imaging on installed CT systems