Focus stacking single-event particle radiography for high spatial resolution images and 3D feature localization

Area of Science:

• Medical Imaging
• Particle Physics
• Computational Imaging

Background:

• Single-event particle radiography (pRad) offers unique imaging capabilities.
• Current pRad techniques face limitations in spatial resolution and 3D feature localization.
• Focus stacking, a technique from optical imaging, has not been applied to pRad.

Purpose of the Study:

• To introduce and validate a novel focus stacking technique for single-event particle radiography (pRad).
• To improve the spatial resolution of pRad images.
• To exploit focus stacking for 3D feature detection and depth localization in pRad.

Main Methods:

• Applied focus stacking to pRad data by combining images reconstructed at different depths using distance-driven binning (DDB).
• Each pixel in the final image is selected based on the maximum gradient from the corresponding pixel in the series of reconstructed images.
• Validated the technique using Geant4 simulations of a water phantom with bone inserts and a lung cancer patient.

Main Results:

• Focus stacking improved median resolution by 136% compared to maximum likelihood reconstruction and 28% compared to standard DDB for a bone phantom.
• Visual resolution improvement was observed in a lung cancer patient's pRad without accuracy loss.
• Depth estimation of features was achieved with millimeter accuracy, with one exception.

Conclusions:

• Focus stacking effectively utilizes inherent 3D information in pRad, overcoming spatial resolution limits.
• The technique enables 3D feature localization, enhancing diagnostic capabilities.
• Focus stacking shows significant potential for advancing future pRad applications in medical imaging.