Human-organ-scale x-ray fluorescence ghost imaging for radioisotope-free diagnostics

A wide range of diagnostic information in medicine is currently obtained using radioactive tracers. While central to nuclear medicine, these methods face inherent limitations: radiation dose restricts repeat examinations, short tracer half-lives and complex logistics limit accessibility and increase costs, and relatively low spatial resolution often necessitates complementary CT or MRI. Here we present a first proof-of-concept demonstration of an alternative imaging modality based on x-ray fluorescence (XRF) computational ghost imaging (CGI) at the human-organ scale. Using a thyroid phantom filled with iodine solution as a model system, we show that structured illuminations combined with fluorescence detection reconstruct the iodine distribution with high fidelity. This approach avoids the use of radioactive tracers while preserving key aspects of the image structure and contrast. We report a spatial resolution of ~ 4.5 mm in the current implementation, with potential for further improvement through system optimization. Beyond this demonstration, XRF-CGI establishes a generalizable framework for tracer imaging without radionuclides, opening a route toward safer, repeatable, and more accessible diagnostics.