Cross-Anatomy Transfer Learning via Shape- Aware Adaptive Fine-Tuning for 3D Vessel Segmentation.

Deep learning methods have recently achieved remarkable performance in vessel segmentation applications, yet require numerous labor-intensive labeled data. To alleviate the requirement of manual annotation, transfer learning methods can potentially be used to acquire the related knowledge of tubular structures from public large-scale labeled vessel datasets for target vessel segmentation in other anatomic sites of the human body. However, the cross-anatomy domain shift is a challenging task due to the formidable discrepancy among various vessel structures in different anatomies, resulting in the limited performance of transfer learning. Therefore, we propose a cross-anatomy transfer learning framework for 3D vessel segmentation, which first generates a pre-trained model on a public hepatic vessel dataset and then adaptively fine-tunes our target segmentation network initialized from the model for segmentation of other anatomic vessels. In the framework, the adaptive fine-tuning strategy is presented to dynamically decide on the frozen or fine-tuned filters of the target network for each input sample with a proxy network. Moreover, we develop a Gaussian-based signed distance map that explicitly encodes vessel-specific shape context. The prediction of the map is added as an auxiliary task in the segmentation network to capture geometry-aware knowledge in the fine-tuning. We demonstrate the effectiveness of our method through extensive experiments on two small-scale datasets of coronary artery and brain vessel. The results indicate the proposed method effectively overcomes the discrepancy of cross-anatomy domain shift to achieve accurate vessel segmentation for these two datasets.

Profiling dendritic cell subsets in the patients with active pulmonary tuberculosis.

Dendritic cell (DC) plays an important role in the immune response against pulmonary tuberculosis. However, the phenotypic profile of DC subsets in peripheral blood in individuals with active pulmonary tuberculosis (APT) is still inconclusive. Here, we demonstrated that the absolute numbers of total DC (tDC), myeloid DC (mDC) and plasmacytoid DC (pDC) in individuals with APT were decreased compared to healthy controls (HCs). The decreased number of DCs, especially of pDC, seems to be a useful diagnostic marker of APT. Meanwhile, the number of DCs was associated with the prolonged/complicated TB, ATD treatment effect and lymphocyte immune reactions, as manifested that relapsed APT patients with a higher number of tDC and lower number of pDC compared to newly diagnosed patients. Interestingly, mDC from APT patients displayed high expressions of CD83 and CCR7, but pDC displayed low expressions of CD83 and CCR7. Moreover, DCs from APT patients expressed lower levels of HLA-DR and CD80, but expressed a higher level of CD86 than those from HCs. However, the antigen uptake capacity of DC subsets was not different between APT and HCs, despite the antigen uptake capacity of pDC was much lower than that of mDC in both APT patients and HCs. Our data represent a systematic profile of DC subsets in the blood of APT patients, and would represent a useful biomarker for APT.

Development and characterization of monoclonal antibody against human IL-37b.

IL-37 has been described as a natural inhibitor of immune responses. Monoclonal antibody (mAb) against human IL-37b with high affinity and specificity can serve as a molecular probe to detect IL-37 and study IL-37 functions, mechanisms and related signal pathways in inflammatory diseases. However, there are very few such mAbs against human IL-37 commercially available so far. In the current study, monoclonal antibodies against human IL-37b were developed by fusing splenocytes from immunized mouse with SP2/0 myeloma cells and polyethylene glycol. Then the antibodies were screened with prokaryotic expressed human IL-37b protein and eukaryotic expressed human IL-37b protein subsequently. Western blot and flow cytometry analysis revealed that selected mAb clons were able to recognize human IL-37 with high specificity. And more importantly, the IL-37b mAbs were fluorescently labeled and can be directly used in flow cytometry and immunohistochemistry. In conclusion, the current study developed new mAbs against human IL-37b, which are applicable in flow cytometry and immunohistochemistry.

Association of CYP1A1 polymorphisms with prostate cancer risk: an updated meta-analysis.

Epidemiological studies have evaluated the association between 3801T>C and 2455A>G polymorphisms of cytochrome P450 1A1 (CYP1A1) and prostate cancer risk. However, controversy exists regarding the role of these polymorphisms. In this work, a meta-analysis was performed to derive a more precise estimation of the relationship. PubMed and ISI Web databases were searched for all cases dated until March 2012. Crude odds ratios with 95 % confidence intervals were used to assess the strength of the association between CYP1A1 polymorphisms and prostate cancer risk. Sensitivity analysis, excluding the studies that deviated from the Hardy-Weinberg equilibrium (HWE), was performed. A total of 17 studies fulfilled our inclusion criteria in this meta-analysis, 12 of which were eligible (1,645 cases and 1,801 controls) for 3801T>C, and eleven (1,640 cases and 1,959 controls) were eligible for 2455A>G. Overall, the 2455A>G polymorphism resulted in a significantly increased susceptibility to prostate cancer. In addition, no significant associations between 3801T>C polymorphism and prostate cancer susceptibility were found in all genetic models. Only an elevated risk was observed for TC versus CC in Asian studies. However, no relationship was found in the Asian group for TC versus CC after excluding the studies that deviated from HWE. Thus, this meta-analysis finds the 2455A>G allele to be a risk factor for prostate cancer, whereas the 3801T>C status does not seem to be capable of modifying prostate cancer risk.

Homozygous intronic mutation leading to inefficient transcription combined with a novel frameshift mutation in F13A1 gene causes FXIII deficiency.

Two novel mutations, 602-605delAAAG in exon 5 and Int1(+12)C>A, of the F13A1 gene were identified in a Chinese factor XIII (FXIII)-deficient family. The 602-605delAAAG mutation results in the premature termination of translation. To determine the functional effect of the Int1(+12)A mutation, we transiently expressed luciferase reporters in U937 cells. We found that the first 951 bp of F13A1 intron 1 is involved in regulating the expression of the F13A1 gene and that Int1(+12)A results in its reduced expression. Electrophoretic mobility shift assay indicated that Int1(+12)A causes reduced protein binding. An Sp1 site was predicted in the sequence containing Int1(+12)C, which the Int1(+12)A mutation eliminates. Co-transfection of a plasmid expressing Sp1 revealed that Sp1 is involved in regulating the expression of FXIIIA and that Int1(+12)A leads to inefficient transcription. These results provide the first insight into a novel regulatory mechanism involving intron 1 in the F13A1 gene.