In this paper, we propose a novel embedding model, named ConvKB, for knowledge base completion. Our model ConvKB advances state-of-the-art models by employing a convolutional neural network, so that it can capture global relationships and transitional characteristics between entities and relations in knowledge bases. In ConvKB, each triple (head entity, relation, tail entity) is represented as a 3column matrix where each column vector represents a triple element. This 3-column matrix is then fed to a convolution layer where multiple filters are operated on the matrix to generate different feature maps. These feature maps are then concatenated into a single feature vector representing the input triple. The feature vector is multiplied with a weight vector via a dot product to return a score. This score is then used to predict whether the triple is valid or not. Experiments show that ConvKB achieves better link prediction performance than previous state-of-the-art embedding models on two benchmark datasets WN18RR and FB15k-237.

Exosomes are nano-scale and closed membrane vesicles which are promising for therapeutic applications due to exosome-enclosed therapeutic molecules such as DNA, small RNAs, proteins and lipids. Recently, it has been demonstrated that mesenchymal stem cell (MSC)-derived exosomes have capacity to regulate many biological events associated with wound healing process, such as cell proliferation, cell migration and blood vessel formation. This study investigated the regenerative potentials for cutaneous tissue, in regard to growth factors associated with wound healing and skin cell proliferation and migration, by exosomes released from primary MSCs originated from bone marrow (BM), adipose tissue (AD), and umbilical cord (UC) under serum-and xeno-free condition. We found crucial wound healing-mediated growth factors, such as vascular endothelial growth factor A (VEGF-A), fibroblast growth factor 2 (FGF-2), hepatocyte growth factor (HGF), and platelet-derived growth factor BB (PDGF-BB) in exosomes derived from all three MSC sources. However, expression levels of these growth factors in exosomes were influenced by MSC origins, especially transforming growth factor beta (TGF-β) was only detected in UCMSC-derived exosomes. All exosomes released by three MSCs sources induced keratinocyte and fibroblast proliferation and migration; and, the induction of cell migration is a dependent manner with the higher dose of exosomes was used (20 µg), the faster migration rate was observed. Additionally, the influences of exosomes on cell proliferation and migration was associated with exosome origins and also target cells of exosomes that the greatest induction of primary dermal fibroblasts belongs to BMMSC-derived exosomes Hoang et al. Mesenchymal Stem Cell-Derived Exosomes and keratinocytes belongs to UCMSC-derived exosomes. Data from this study indicated that BMMSCs and UCMSCs under clinical condition secreted exosomes are promising to develop into therapeutic products for wound healing treatment.

Electronic medical record (EMR) offers promises for novel analytics. However, manual feature engineering from EMR is labor intensive because EMR is complex - it contains temporal, mixed-type and multimodal data packed in irregular episodes. We present a computational framework to harness EMR with minimal human supervision via restricted Boltzmann machine (RBM). The framework derives a new representation of medical objects by embedding them in a low-dimensional vector space. This new representation facilitates algebraic and statistical manipulations such as projection onto 2D plane (thereby offering intuitive visualization), object grouping (hence enabling automated phenotyping), and risk stratification. To enhance model interpretability, we introduced two constraints into model parameters: (a) nonnegative coefficients, and (b) structural smoothness. These result in a novel model called eNRBM (EMR-driven nonnegative RBM). We demonstrate the capability of the eNRBM on a cohort of 7578 mental health patients under suicide risk assessment. The derived representation not only shows clinically meaningful feature grouping but also facilitates short-term risk stratification. The F-scores, 0.21 for moderate-risk and 0.36 for high-risk, are significantly higher than those obtained by clinicians and competitive with the results obtained by support vector machines.

No abstract

In this paper, we introduce an embedding model, named CapsE, exploring a capsule network to model relationship triples (subject, relation, object). Our CapsE represents each triple as a 3-column matrix where each column vector represents the embedding of an element in the triple. This 3-column matrix is then fed to a convolution layer where multiple filters are operated to generate different feature maps. These feature maps are reconstructed into corresponding capsules which are then routed to another capsule to produce a continuous vector. The length of this vector is used to measure the plausibility score of the triple. Our proposed CapsE obtains better performance than previous state-of-the-art embedding models for knowledge graph completion on two benchmark datasets WN18RR and FB15k-237, and outperforms strong search personalization baselines on SEARCH17.

Background aims: Mesenchymal stem/stromal cells (MSCs) are of interest for the treatment of graft-versushost disease, autoimmune diseases, osteoarthritis and neurological and cardiovascular diseases. Increasing numbers of clinical trials emphasize the need for standardized manufacturing of these cells. However, many challenges related to diverse isolation and expansion protocols and differences in cell tissue sources exist. As a result, the cell products used in numerous trials vary greatly in characteristics and potency. Methods: The authors have established a standardized culture platform using xeno-and serum-free commercial media for expansion of MSCs derived from umbilical cord (UC), bone marrow and adipose-derived (AD) and examined their functional characteristics. Results: MSCs from the tested sources stably expanded in vitro and retained their biomarker expression and normal karyotype at early and later passages and after cryopreservation. MSCs were capable of colony formation and successfully differentiated into osteogenic, adipogenic and chondrogenic lineages. Pilot expansion of UC-MSCs and AD-MSCs to clinical scale revealed that the cells met the required quality standard for therapeutic applications. Conclusions: The authors' data suggest that xeno-and serum-free culture conditions are suitable for largescale expansion and enable comparative study of MSCs of different origins. This is of importance for therapeutic purposes, especially because of the numerous variations in pre-clinical and clinical protocols for MSCbased products.

Detecting anomalies in surveillance videos has long been an important but unsolved problem. In particular, many existing solutions are overly sensitive to (often ephemeral) visual artifacts in the raw video data, resulting in false positives and fragmented detection regions. To overcome such sensitivity and to capture true anomalies with semantic significance, one natural idea is to seek validation from abstract representations of the videos. This paper introduces a framework of robust anomaly detection using multilevel representations of both intensity and motion data. The framework consists of three main components: 1) representation learning using Denoising Autoencoders, 2) level-wise representation generation using Conditional Generative Adversarial Networks, and 3) consolidating anomalous regions detected at each representation level. Our proposed multilevel detector shows a significant improvement in pixel-level Equal Error Rate, namely 11.35%, 12.32% and 4.31% improvement in UCSD Ped 1, UCSD Ped 2 and Avenue datasets respectively. In addition, the model allowed us to detect mislabeled anomalies in the UCDS Ped 1.

No abstract