This study introduces the qualitative and quantitative characteristics of x-ray dark-field chest radiographs in healthy humans. Background X-ray dark-field radiography takes advantage of the wave properties of x-rays, with a relatively high signal in the lungs due to the many air-tissue interfaces in the alveoli. Purpose To...

Dark-field chest radiography shows dark-field signal intensity losses that correspond to emphysematous areas found on chest CT images in participants with emphysema. Background Dark-field chest radiography allows for assessment of lung alveolar structure by exploiting wave optical properties of x-rays. Purpose To evaluate the qualitative and quantitative featur...

To evaluate the diagnostic value of MR imaging for the differentiation of lipomas and atypical lipomatous tumors (ALT) in comparison with histology and MDM2 amplification status. Patients with well-differentiated lipomatous tumors (n = 113), of which 66 were diagnosed as lipoma (mean age 53 years (range, 13–82); 47% women) and 47 as atypical lipomatous tumor (ALT; mean age 60 years (range, 28–88); 64% women), were included into this study using histology and MDM2 amplification status by fluorescence in situ hybridization (FISH) as standard of reference. Preoperative MR images were retrospectively assessed by two radiologists for the following imaging features: maximum tumor diameter (mm) as well as the affected compartment (intramuscular, intermuscular or subcutaneous), septa (absent, thin (< 2 mm) or thick septa (> 2 mm) with nodular components); contrast enhancing areas within the lipomatous tumor (< 1/3 of the tumor volume, > 1/3 of the tumor volume); Of the 47 patients with ALT, 40 (85.1%) presented thick septa (> 2 mm) and this finding significantly increased the likelihood of ALT (OR 6.24, 95% CI 3.36–11.59; P < 0.001). The likelihood of ALT was increased if the tumor exceeded a maximum diameter of 130.0 mm (OR 2.74, 95% CI 1.82–4.11, P < 0.001). The presence of contrast enhancement in lipomatous tumors significantly increased the likelihood of ALT (Odds ratio (OR) 2.95, 95% confidence interval (CI) 2.01–4.31; P < 0.001). Of the lipomas, 21.1% were located subcutaneously, 63.6% intramuscularly and 15.2% intermuscularly. On the other hand, none of the ALTs were located subcutaneously, the majority was located intermuscularly (87.3%) and a small number of ALTs was located intramuscularly (12.7%). Our results suggest that using specific morphological MR imaging characteristics (maximum tumor diameter, thick septa and contrast enhancement) and the information on the localization of the lipomatous tumor, a high sensitivity and substantial specificity can be achieved for the diagnosis of lipomas and ALTs.

Deep learning based solutions are being succesfully implemented for a wide variety of applications. Most notably, clinical use-cases have gained an increased interest and have been the main driver behind some of the cutting-edge data-driven algorithms proposed in the last years. For applications like sparse-view tomographic reconstructions, where the amount of measurement data is small in order to keep acquisition time short and radiation dose low, reduction of the streaking artifacts has prompted the development of data-driven denoising algorithms with the main goal of obtaining diagnostically viable images with only a subset of a full-scan data. We propose WNet, a data-driven dual-domain denoising model which contains a trainable reconstruction layer for sparse-view artifact denoising. Two encoder-decoder networks perform denoising in both sinogram- and reconstruction-domain simultaneously, while a third layer implementing the Filtered Backprojection algorithm is sandwiched between the first two and takes care of the reconstruction operation. We investigate the performance of the network on sparse-view chest CT scans, and we highlight the added benefit of having a trainable reconstruction layer over the more conventional fixed ones. We train and test our network on two clinically relevant datasets and we compare the obtained results with three different types of sparse-view CT denoising and reconstruction algorithms.