Search

Given their powerful feature representation for recognition, deep convolutional neural networks (DCNNs) have been driving rapid advances in high-level computer vision tasks. However, their performance in semantic image segmentation is still not satisfactory. Based on the analysis of visual mechanism, we conclude that DCNNs in a bottom-up manner are not enough, because semantic image segmentation task requires not only recognition but also visual attention capability. In the study, superpixels containing visual attention information are introduced in a top-down manner, and an extensible architecture is proposed to improve the segmentation results of current DCNN-based methods. We employ the current state-of-the-art fully convolutional network (FCN) and FCN with conditional random field (DeepLab-CRF) as baselines to validate our architecture. Experimental results of the PASCAL VOC segmentation task qualitatively show that coarse edges and error segmentation results are well improved. We also quantitatively obtain about 2%-3% intersection over union (IOU) accuracy improvement on the PASCAL VOC 2011 and 2012 test sets.

Total generalized variation (TGV)-based computed tomography (CT) image reconstruction, which utilizes high-order image derivatives, is superior to total variation-based methods in terms of the preservation of edge information and the suppression of unfavorable staircase effects. However, conventional TGV regularization employs l1-based form, which is not the most direct method for maximizing sparsity prior. In this study, we propose a total generalized p-variation (TGpV) regularization model to improve the sparsity exploitation of TGV and offer efficient solutions to few-view CT image reconstruction problems. To solve the nonconvex optimization problem of the TGpV minimization model, we then present an efficient iterative algorithm based on the alternating minimization of augmented Lagrangian function. All of the resulting subproblems decoupled by variable splitting admit explicit solutions by applying alternating minimization method and generalized p-shrinkage mapping. In addition, approximate solutions that can be easily performed and quickly calculated through fast Fourier transform are derived using the proximal point method to reduce the cost of inner subproblems. The accuracy and efficiency of the simulated and real data are qualitatively and quantitatively evaluated to validate the efficiency and feasibility of the proposed method. Overall, the proposed method exhibits reasonable performance and outperforms the original TGV-based method when applied to few-view problems.

Stem cells are characterized by their capability to self-renew and terminally differentiate into multiple cell types. Somatic or adult stem cells have a finite self-renewal capacity and are lineage-restricted. The use of adult stem cells for therapeutic purposes has been a topic of recent interest given the ethical considerations associated with embryonic stem (ES) cells. Mesenchymal stem cells (MSCs) are adult stem cells that can differentiate into osteogenic, adipogenic, chondrogenic, or myogenic lineages. Owing to their ease of isolation and unique characteristics, MSCs have been widely regarded as potential candidates for tissue engineering and repair. While various signaling molecules important to MSC differentiation have been identified, our complete understanding of this process is lacking. Recent investigations focused on the role of epigenetic regulation in lineage-specific differentiation of MSCs have shown that unique patterns of DNA methylation and histone modifications play an important role in the induction of MSC differentiation toward specific lineages. Nevertheless, MSC epigenetic profiles reflect a more restricted differentiation potential as compared to ES cells. Here we review the effect of epigenetic modifications on MSC multipotency and differentiation, with a focus on osteogenic and adipogenic differentiation. We also highlight clinical applications of MSC epigenetics and nuclear reprogramming.

Osteosarcoma is the most common primary malignancy of bone with poorly characterized molecular pathways important in its pathogenesis. Increasing evidence indicates that elevated lipid biosynthesis is a characteristic feature of cancer. We sought to investigate the role of lysophosphatidic acid acyltransferase β (LPAATβ, aka, AGPAT2) in regulating the proliferation and growth of human osteosarcoma cells. LPAATβ can generate phosphatidic acid, which plays a key role in lipid biosynthesis as well as in cell proliferation and survival. Although elevated expression of LPAATβ has been reported in several types of human tumors, the role of LPAATβ in osteosarcoma progression has yet to be elucidated.Endogenous expression of LPAATβ in osteosarcoma cell lines is analyzed by using semi-quantitative PCR and immunohistochemical staining. Adenovirus-mediated overexpression of LPAATβ and silencing LPAATβ expression is employed to determine the effect of LPAATβ on osteosarcoma cell proliferation and migration in vitro and osteosarcoma tumor growth in vivo. We have found that expression of LPAATβ is readily detected in 8 of the 10 analyzed human osteosarcoma lines. Exogenous expression of LPAATβ promotes osteosarcoma cell proliferation and migration, while silencing LPAATβ expression inhibits these cellular characteristics. We further demonstrate that exogenous expression of LPAATβ effectively promotes tumor growth, while knockdown of LPAATβ expression inhibits tumor growth in an orthotopic xenograft model of human osteosarcoma.Our results strongly suggest that LPAATβ expression may be associated with the aggressive phenotypes of human osteosarcoma and that LPAATβ may play an important role in regulating osteosarcoma cell proliferation and tumor growth. Thus, targeting LPAATβ may be exploited as a novel therapeutic strategy for the clinical management of osteosarcoma. This is especially attractive given the availability of selective pharmacological inhibitors.