Graph Convolutional Network Based on Manifold Similarity Learning

In the area of large-scale graph data representation and semi-supervised learning, deep graph-based convolutional neural networks have been widely applied. However, typical graph convolutional network (GCN) aggregates information of neighbor nodes based on binary neighborhood similarity (adjacency matrix). It treats all neighbor nodes of one node equally, which does not suppress the influence of dissimilar neighbor nodes. In this paper, we investigate GCN based on similarity matrix instead of adjacency matrix of graph nodes. Gaussian heat kernel similarity in Euclidean space is first adopted, which is named EGCN. Then biologically inspired manifold similarity is trained in reproducing kernel Hilbert space (RKHS), based on which a manifold GCN (named MGCN) is proposed for graph data representation and semi-supervised learning with four different kernel types. The proposed method is evaluated with extensive experiments on four benchmark document citation network datasets. The objective function of manifold similarity learning converges very quickly on different datasets using various kernel functions. Compared with state-of-the-art methods, our method is very competitive in terms of graph node recognition accuracy. In particular, the recognition rates of MGCN (Gaussian kernel) and MGCN (Polynomial Kernel) outperform that of typical GCN about 3.8% on Cora dataset, 3.5% on Citeseer dataset, 1.3% on Pubmed dataset and 4% on Cora_ML dataset, respectively. Although the proposed MGCN is relatively simple and easy to implement, it can discover local manifold structure by manifold similarity learning and suppress the influence of dissimilar neighbor nodes, which shows the effectiveness of the proposed MGCN.

     

Prostate thermal therapy with high intensity transurethral ultrasound: the impact of pelvic bone heating on treatment delivery.

PURPOSE: This study was designed to assess pelvic bone temperature during typical treatment regimens of transurethral ultrasound thermal ablation of the prostate to establish guidelines for limiting bone heating. METHODS: Treatment with transurethral planar, curvilinear, and sectored tubular applicators was simulated using an acoustic and biothermal pelvic model that accommodates applicator sweeping, boundary temperature control, and changes in perfusion and attenuation with thermal dose to more accurately model ultrasound energy penetration. The effects of various parameters including power and frequency (5-10 MHz) on bone heating were assessed for a range of prostate cross-sections (3-5 cm) and bone distances (1-3 cm). RESULTS: All devices can produce significant bone heating (temperatures >50 degrees C, thermal dose >240 EM(43 degrees C)) without optimization of applied frequency or power for bone <3 cm from the prostate boundary. In small glands ( approximately 3 cm) increasing operating frequency of curvilinear and planar devices can increase bone temperatures, whereas the tubular applicator can be used at 10 MHz to avoid likely bone damage. In larger prostates (4-5 cm wide) increasing frequency reduces bone heating but can substantially increase treatment time. Lowering power can reduce bone temperature but may increase thermal dose by increasing treatment duration. All applicators can be used to treat glands 4-5 cm with limited bone heating by selecting appropriate power and frequency. CONCLUSIONS: Pubic bone heating during ultrasound thermal therapy of the prostate can be substantial in certain situations. Successful realization of this therapy will require patient-specific treatment planning to optimally determine power and frequency in order to minimize bone heating.     

Infection control issues in central venous catheter care.

Central venous catheters (CVCs) are now a routine part of patient management in the intensive care unit (ICU). Over time, a vast amount of literature associated with the use and care of CVCs has accumulated. The purpose of this article is to discuss the literature associated with the care of these devices in a narrative format. Although particular attention is paid to infection control issues, other fundamental areas such as catheter design, dressings, line changing and post insertion management are also discussed. The article goes on to look at the future of CVC design and concludes with an analysis of future developments related to CVCs.