Labeled medical imaging data is scarce and expensive to generate. To achieve generalizable deep learning models large amounts of data are needed. Standard data augmentation is a method to increase generalizability and is routinely performed. Generative adversarial networks offer a novel method for data augmentation. We evaluate the use of CycleGAN for data augmentation in CT segmentation tasks. Using a large image database we trained a CycleGAN to transform contrast CT images into non-contrast images. We then used the trained CycleGAN to augment our training using these synthetic non-contrast images. We compared the segmentation performance of a U-Net trained on the original dataset compared to a U-Net trained on the combined dataset of original data and synthetic non-contrast images. We further evaluated the U-Net segmentation performance on two separate datasets: The original contrast CT dataset on which segmentations were created and a second dataset from a different hospital containing only non-contrast CTs. We refer to these 2 separate datasets as the in-distribution and out-of-distribution datasets, respectively. We show that in several CT segmentation tasks performance is improved significantly, especially in out-of-distribution (noncontrast CT) data. For example, when training the model with standard augmentation techniques, performance of segmentation of the kidneys on out-of-distribution non-contrast images was dramatically lower than for in-distribution data (Dice score of 0.09 vs. 0.94 for out-of-distribution vs. in-distribution data, respectively, p < 0.001). When the kidney model was trained with CycleGAN augmentation techniques, the out-of-distribution (non-contrast) performance increased dramatically (from a Dice score of 0.09 to 0.66, p < 0.001). Improvements for the liver and spleen were smaller, from 0.86 to 0.89 and 0.65 to 0.69, respectively. We believe this method will be valuable to medical imaging researchers to reduce manual segmentation effort and cost in CT imaging.

Integrin signaling critically contributes to the progression, growth, and therapy resistance of malignant tumors. Here, we show that targeting of β 1 integrins with inhibitory antibodies enhances the sensitivity to ionizing radiation and delays the growth of human head and neck squamous cell carcinoma cell lines in 3D cell culture and in xenografted mice. Mechanistically, dephosphorylation of focal adhesion kinase (FAK) upon inhibition of β 1 integrin resulted in dissociation of a FAK/cortactin protein complex. This, in turn, downregulated JNK signaling and induced cell rounding, leading to radiosensitization. Thus, these findings suggest that robust and selective pharmacological targeting of β 1 integrins may provide therapeutic benefit to overcome tumor cell resistance to radiotherapy.

The use of biologic therapy has increased over the past decade well beyond primary autoimmune diseases. Indeed, a recent trial using an anti-IL-1beta antibody reduced second myocardial infarction (MI) in those who have had MI. Psoriasis is a chronic inflammatory disease often treated with biologics when severe, is associated with increased risk of MI, in part driven by high-risk coronary plaque phenotypes by coronary computed tomography angiography (CCTA). We hypothesized that we would observe a reduction in inflammatory-driven phenotypes of coronary plaque, including non-calcified coronary plaque burden and lipid-rich necrotic core in those treated with biologic therapy after one-year compared with non-biologic therapy.

Background Psoriasis, a chronic inflammatory disease associated with an accelerated risk of MI, provides an ideal human model to study inflammatory atherogenesis in vivo. We hypothesized that the increased cardiovascular risk observed in psoriasis would be partially attributable to an elevated subclinical coronary artery disease (CAD) burden composed of non-calcified plaques with high-risk features. However, inadequate efforts have been made to directly measure CAD in this vulnerable population. As such, we sought to compare total (TB) and non-calcified (NCB) coronary plaque burden, and high-risk plaque (HRP) prevalence, between psoriasis patients (n=105), hyperlipidemic patients eligible for statin therapy under NCEP-ATP III guidelines (n=100) who were ~10 years older, and non-psoriasis healthy volunteers (HV) (n=25). Methods Patients underwent coronary computed-tomography angiography (CCTA) for TB and NCB quantification, and HRP identification, defined as low-attenuation (<30 HU), positive remodeling (>1.10), and spotty calcification. A consecutive sample of the first 50 psoriasis patients were scanned again at 1 year following therapy. Results Despite being younger and at lower traditional risk than hyperlipidemic patients, psoriasis patients had increased NCB (mean±S.D.:1.18±0.33 vs 1.11±0.32, p=0.02), and similar HRP prevalence (p=0.58). Furthermore, compared to HV, psoriasis patients had increased TB (1.22±0.31 vs 1.04±0.22, p=0.001), NCB (1.18±0.33 vs 1.03±0.21, p=0.004), and HRP prevalence beyond traditional risk (OR=6.0, 95% CI: 1.1–31.7; p=0.03). Finally, amongst psoriasis patients followed for 1-year, improvement in psoriasis severity associated with improvement in TB (β=0.45, 0.23–0.67; p<0.001) and NCB (β=0.53, 0.32–0.74; p<0.001) beyond traditional risk factors. Conclusions Psoriasis patients had greater NCB and increased HRP prevalence than HV. Additionally, psoriasis patients had elevated NCB and equivalent HRP prevalence as older, hyperlipidemic patients. Finally, modulation of target organ inflammation (eg. skin) associated with an improvement in NCB at 1 year, suggesting that control of remote sites of inflammation may translate into reduced CAD risk.