Search

Abstract The separation of mixtures of substances into their individual components plays an important role in many areas of science. In medical imaging, one method is the established analysis using dual-energy computed tomography. However, when analyzing mixtures consisting of more than three individual basis materials, a physical limit is reached that no longer allows this standard analysis. In addition, the X-ray attenuation coefficients of chemically complicated basis materials may not be known and also cannot be determined by other or previous analyses. To address these issues, we developed a novel theoretical approach and algorithm and tested it on samples prepared in the laboratory as well as on ex-vivo medical samples. This method allowed both five-material decomposition and determination or optimization of the X-ray attenuation coefficients of the sample base materials via optimizations of objective functions. After implementation, this new multimodal method was successfully tested on self-mixed samples consisting of the aqueous base solutions iomeprol, eosin Y disodiumsalt, sodium chloride, and pure water. As a first proof of concept of this technique for detailed material decomposition in medicine we analyzed exact percentage composition of ex vivo clots from patients with acute ischemic stroke, using histological analysis as a reference standard.

Abstract Gout is the most common form of inflammatory arthritis, caused by the deposition of monosodium urate (MSU) crystals in peripheral joints and tissue. Detection of MSU crystals is essential for definitive diagnosis, however the gold standard is an invasive process which is rarely utilized. In fact, most patients are diagnosed or even misdiagnosed based on manifested clinical signs, as indicated by the unchanged premature mortality among gout patients over the past decade, although effective treatment is now available. An alternative, non-invasive approach for the detection of MSU crystals is X-ray dark-field radiography. In our work, we demonstrate that dark-field X-ray radiography can detect naturally developed gout in animals with high diagnostic sensitivity and specificity based on the in situ measurement of MSU crystals. With the results of this study as a potential basis for further research, we believe that X-ray dark-field radiography has the potential to substantially improve gout diagnostics.

Abstract Grating-based phase-contrast computed tomography (GBPC-CT) enables increased soft tissue differentiation, but often suffers from streak artifacts when performing high-sensitivity GBPC-CT of biomedical samples. Current GBPC-CT setups consist of one-dimensional gratings and hence allow to measure only the differential phase-contrast (DPC) signal perpendicular to the direction of the grating lines. Having access to the full two-dimensional DPC signal can strongly reduce streak artefacts showing up as characteristic horizontal lines in the reconstructed images. GBPC-CT with gratings tilted by 45° around the optical axis, combining opposed projections, and reconstructing with filtered backprojection is one method to retrieve the full three-dimensional DPC signal. This approach improves the quality of the tomographic data as already demonstrated at a synchrotron facility. However, additional processing and interpolation is necessary, and the approach fails when dealing with cone-beam geometry setups. In this work, we employ the tilted grating configuration with a laboratory GBPC-CT setup with cone-beam geometry and use statistical iterative reconstruction (SIR) with a forward model accounting for diagonal grating alignment. Our results show a strong reduction of streak artefacts and significant increase in image quality. In contrast to the prior approach our proposed method can be used in a laboratory environment due to its cone-beam compatibility.