Dark-Field X-Ray Imaging Significantly Improves Deep-Learning based Detection of Synthetic Early-Stage Lung Tumors in Preclinical Models

Low-dose computed tomography (LDCT) is the current standard for lung cancer screening, yet its adoption and accessibility remain limited. Many regions lack LDCT infrastructure, and even among those screened, early-stage cancer detection often yield false positives, as shown in the National Lung Screening Trial (NLST) with a sensitivity of 93.8 percent and a false-positive rate of 26.6 percent. We aim to investigate whether X-ray dark-field imaging (DFI) radiograph, a technique sensitive to small-angle scatter from alveolar microstructure and less susceptible to organ shadowing, can significantly improve early-stage lung tumor detection when coupled with deep-learning segmentation. Using paired attenuation (ATTN) and DFI radiograph images of euthanized mouse lungs, we generated realistic synthetic tumors with irregular boundaries and intensity profiles consistent with physical lung contrast. A U-Net segmentation network was trained on small patches using either ATTN, DFI, or a combination of ATTN and DFI channels. Results show that the DFI-only model achieved a true-positive detection rate of 83.7 percent, compared with 51 percent for ATTN-only, while maintaining comparable specificity (90.5 versus 92.9 percent). The combined ATTN and DFI input achieved 79.6 percent sensitivity and 97.6 percent specificity. In conclusion, DFI substantially improves early-tumor detectability in comparison to standard attenuation radiography and shows potential as an accessible, low-cost, low-dose alternative for pre-clinical or limited-resource screening where LDCT is unavailable.

Reducing Artifacts in Grating Interferometry Using Multiple Harmonics and Phase Step Corrections

X-ray interferometry is an emerging imaging modality with a wide variety of potential clinical applications, including lung and breast imaging, as well as in non-destructive testing, such as additive manufacturing and porosimetry. A grating interferometer uses a diffraction grating to produce a periodic interference pattern and measures how a patient or sample perturbs the pattern, producing three unique images that highlight X-ray absorption, refraction, and small angle scattering, known as the transmission, differential-phase, and dark-field images, respectively. Image artifacts that are unique to X-ray interferometry are introduced when assuming the fringe pattern is perfectly sinusoidal and the phase steps are evenly spaced. Inaccuracies in grating position, coupled with multi-harmonic fringes, lead to remnant oscillations and phase wraparound artifacts. We have developed an image recovery algorithm that uses additional harmonics, direct relative phase fitting, and phase step corrections to prevent them. The direct relative phase fitting removes the phase wraparound artifact. Correcting the phase step positions and introducing the additional harmonic removes the grating remnant artifact present in the transmission, differential-phase, and dark-field images. By modifying existing algorithms, the fit to the fringe pattern is greatly improved and artifacts are minimized, as we demonstrate with the imaging of several samples, including PMMA microspheres, ex vivo formalin fixed mouse lungs, and porous alumina.