Sub pixel resolution using spectral-spatial encoding in x-ray imaging.

Area of Science:

• Medical Imaging
• X-ray Technology
• Image Resolution Enhancement

Background:

• Traditional methods for improving X-ray spatial resolution involve reducing focal spot or detector element size.
• Many super-resolution techniques necessitate physical movement of imaging system components.
• A need exists for methods that enhance resolution without object or detector motion.

Purpose of the Study:

• To introduce a novel method for achieving spatial resolution beyond the physical limits of detector pixels.
• To demonstrate a technique that enhances spatial resolution without requiring physical movement of the object or detector.

Main Methods:

• Introduced sub-pixel encoding (SPE), a method weighting photon energy spectra on a sub-pixel scale using physical filters.
• Acquired multiple X-ray projections to synthesize information at a spatial scale determined by spectrum modulation.
• Demonstrated SPE feasibility using simulated energy-integrating detectors for medical and industrial applications, resolving features smaller than detector pitch.

Main Results:

• SPE successfully synthesized information at a finer spatial scale than the detector element size.
• Simulations showed phantom features resolvable by SPE with a 2 mm detector that were not visible with a standard 2 mm detector, but were resolvable with a 1 mm detector.
• Covariance matrix analysis characterized noise propagation and correlation, with negative diagonal terms observed in example cases.

Conclusions:

• Introduced the concept of encoding and retrieving object information at a sub-pixel length scale.
• SPE enables simultaneous enhancement of spatial resolution and provides dual-energy-like information on photon-matter interactions.