Single-shot phase-contrast computed tomography for multi-material samples

Abstract

Edge illumination (EI) is a mask-based X-ray phase-contrast imaging (XPCI) technique that converts small refraction-induced beam deviations into measurable intensity changes, enabling laboratory-compatible imaging with sensitivity beyond conventional absorption contrast. However, quantitative EI computed tomography typically requires multiple exposures per projection and computationally demanding iterative reconstruction, increasing acquisition time and radiation dose. Here, we present a single-shot technique for quantitative EI phase-contrast CT of multi-material samples based on a linear joint reconstruction (LJR) approach. The method uses a single exposure per angle of projection and directly reconstructs the sample’s absorption and phase components simultaneously. This is achieved by applying a first-order Taylor expansion to the system’s illumination curve, which linearizes the EI imaging model and transforms it into a convex least-squares problem with a unique semi-analytical solution, eliminating the non-convexity that typically leads to local minima in iterative approaches. As a result, the approach is both fast and robust. We demonstrate the method’s performance on both simulated and experimental datasets, including multi-material plastic rods and an intact mouse brain inside its native skull. The results show that the proposed technique produces high-quality, quantitatively accurate phase and absorption tomograms that are a close match to those obtained with the standard EI method while considerably reducing data acquisition and computational time. This development enables rapid, reliable XPCI of complex multi-material samples under challenging imaging conditions.