Universal Phase Contrast in Micro-CT Systems

Abstract

Conventional high-resolution micro-CT systems are regarded as attenuation-based unless visible Fresnel fringes reveal the presence of propagation-based phase contrast. Here we show that this interpretation is incomplete. When propagation-induced phase transfer is non-negligible relative to the system blur, micro-CT operates in a phase-transfer regime in which propagation improves spatial resolution relative to that expected from source and detector blur alone, irrespective of whether Fresnel fringes remain visible. Because micro-CT systems possess finite source and detector blur, this phase transfer is accompanied by a degree of hardware-induced phase retrieval (HIPR), ranging from under-HIPR to matched-HIPR and over-HIPR. We further identify detector-induced phase retrieval (DIPR) as the optimal case in which the detector provides all of the filtering required for phase retrieval, thereby preserving the phase-induced resolution enhancement while introducing spatial correlations between detected photons that reduce high-spatial-frequency noise. We derive analytical conditions for HIPR and DIPR, introduce frequency-domain metrics quantifying preserved phase transfer and detector compatibility, and validate the framework numerically with simulations and experimentally using custom and commercial micro-CT systems. Our results demonstrate that phase-induced resolution enhancement can occur in conventional micro-CT systems even in the absence of visible Fresnel fringes, changing how image formation, spatial resolution, and optimization should be interpreted.