Demo dataset and NightRail inputs fo the paper: {Eikonal Phase Retrieval with the WKB Diffraction-Pressure Term and a Fast Second-Order Implementation for High-Throughput Beamlines

Propagation-based phase-contrast micro-tomography (PPC‑µCT) in the near-field regime exploits long sample-to-detector distances to enhance weak refractive features while preserving experimental simplicity and dose efficiency. The reduced source size of fourth‑generation storage rings extends the usable propagation range, but in high-throughput beamlines such as ESRF‑BM18 the imaging of large, strongly absorbing specimens combines low‑contrast regions with sharp phase gradients (e.g. at bone/air or bubble interfaces). Under these conditions, the linearized single-distance transport model underpinning widely used Paganin-type filters breaks down, and nonlinear phase‑propagation effects generate characteristic streak artifacts. In particular, tangential rays can produce bright or dark lines when two interfaces become simultaneously tangent and the corresponding phase gradients add or cancel, respectively. Here we present a semi‑analytical eikonal phase‑retrieval framework that retains a second‑order expansion in propagation distance L, providing a fast forward model beyond first‑order (linearized) TIE-based retrieval. The same asymptotic formalism naturally accommodates the leading WKB diffraction‑pressure correction, enabling a controlled wave‑optics refinement of the geometrical transport picture. The resulting second‑order solver improves artifact suppression in mixed‑attenuation objects and reduces computational cost, making nonlinear phase retrieval practical for routine, high‑throughput PPC‑µCT pipelines.
This dataset provides data for the demo and the inputs for NightRail