Search

Background Despite the wide range of cleft lip morphology, consistent scales to categorize preoperative severity do not exist. Machine learning has been used to increase accuracy and efficiency in detection and rating of multiple conditions, yet it has not been applied to cleft disease. The authors tested a machine learning approach to automatically detect and measure facial landmarks and assign severity grades using preoperative photographs. Methods Preoperative images were collected from 800 unilateral cleft lip patients, manually annotated for cleft-specific landmarks, and rated using a previously validated severity scale by eight expert reviewers. Five convolutional neural network models were trained for landmark detection and severity grade assignment. Mean squared error loss and Pearson correlation coefficient for cleft width ratio, nostril width ratio, and severity grade assignment were calculated. Results All five models performed well in landmark detection and severity grade assignment, with the largest and most complex model, Residual Network, performing best (mean squared error, 24.41; cleft width ratio correlation, 0.943; nostril width ratio correlation, 0.879; severity correlation, 0.892). The mobile device-compatible network, MobileNet, also showed a high degree of accuracy (mean squared error, 36.66; cleft width ratio correlation, 0.901; nostril width ratio correlation, 0.705; severity correlation, 0.860). Conclusions Machine learning models demonstrate the ability to accurately measure facial features and assign severity grades according to validated scales. Such models hold promise for the creation of a simple, automated approach to classifying cleft lip morphology. Further potential exists for a mobile telephone-based application to provide real-time feedback to improve clinical decision making and patient counseling.

Nylon 11 is a polymer synthesized from 11-aminoundecanoic acid, and widely used in commercial manufacturing. In this study, we describe the isolation of the first organism capable of metabolizing 11-aminoundecanoic acid from nylon 11 enrichment culture. The strain shows rapid growth on 11-aminoundecanoic acid as a sole source of carbon, nitrogen, and energy. Furthermore, the genome sequence of strain JG-B was deciphered and shown to belong to genus Pseudomonas. Many genes encoding putative extracellular hydrolases, as well as homologues of nylon 6 hydrolases (NylB and NylA) were identified, suggesting the metabolic versatility and possibility that this organism could also depolymerase nylon 11 polymers.

Evidence is emerging that disruption of the endothelial glycocalyx might contribute importantly to arterial dysfunction in the context of diabetes. One approach to assess the integrity of the endothelium and the vascular smooth muscle cell layer, in the absence of neural, humoral, and mechanical influences, is by measuring arterial vasomotion ex vivo. Here we describe a procedure to assess non-receptor-mediated vasoconstriction, receptor-mediated vasoconstriction, and endothelium-dependent and -independent vasodilation, in resistance and conductance arteries pressurized to 60 mmHg. In addition to evaluating vasoreactivity using isobaric approaches, the same experimental set-up can be used to initiate a pressure gradient across the artery such that intraluminal, flow-mediated vasodilation can be measured. After recording endothelium-dependent vasodilation using isobaric or flow-mediated approaches, identical interventions can be completed in the presence of enzymes that cleave biologically active heparan sulfates into inactive disaccharide and oligosaccharide fragments to assess the contribution from: (a) endothelial-derived substances (e.g., nitric oxide via nitric oxide synthase inhibition); or (b) important components of the glycocalyx (e.g., removal of heparan sulfate via heparitinase III treatment). Here, we show that acute disruption of a predominant glycosaminoglycan i.e., heparan sulfate impairs intraluminal flow-mediated vasodilation in murine resistance arteries.

Advances in artificial intelligence and in particular machine learning and neural networks have given rise to a new generation of virtual assistants and chatbots. Within this work, we present NADiA - Neurally Animated Dialog Agent - that leverages both the user's verbal input as well as their facial expressions to respond in a meaningful way. NADiA combines a neural language model that generates appropriate responses to user prompts, a convolutional neural network for facial expression analysis, and virtual human technology that is deployed on a mobile phone. Here, we evaluate NADiA's anthropomorphic characteristics and its ability to understand the human interlocutor using both subjective as well as objective measures. We find that NADiA significantly outperforms state of the art chatbot technology and produces comparable behavior to human generated reference outputs.

Diffusion is commonly used as a ranking or re-ranking method in retrieval tasks to achieve higher retrieval performance, and has attracted lots of attention in recent years. A downside to diffusion is that it performs slowly in comparison to the naive k-NN search, which causes a non-trivial online computational cost on large datasets. To overcome this weakness, we propose a novel diffusion technique in this paper. In our work, instead of applying diffusion to the query, we pre-compute the diffusion results of each element in the database, making the online search a simple linear combination on top of the k-NN search process. Our proposed method becomes 10~ times faster in terms of online search speed. Moreover, we propose to use late truncation instead of early truncation in previous works to achieve better retrieval performance., Comment: Accepted by AAAI 2019

Most marine invertebrates and algae are osmoconformers whose cells accumulate organic osmolytes that provide half or more of cellular osmotic pressure. These solutes are primarily free amino acids and glycine betaine in most invertebrates and small carbohydrates and dimethylsulfoniopropionate (DMSP) in many algae. Corals with endosymbiotic dinoflagellates (Symbiodinium spp.) have been reported to obtain from the symbionts potential organic osmolytes such as glycerol, amino acids, and DMSP. However, corals and their endosymbionts have not been fully analyzed for osmolytes. We quantified small carbohydrates, free amino acids, methylamines, and DMSP in tissues of the corals Fungia scutaria, Pocillopora damicornis, Pocillopora meandrina, Montipora capitata, Porites compressa, and Porites lobata (all with symbionts) plus Tubastrea aurea (asymbiotic) from Kaneohe Bay, Oahu (Hawaii). Glycine betaine, at 33-69 mmol/kg wet mass, was found to constitute 90% or more of the measured organic solutes in all except the Porites species. Those were dominated by proline betaine and dimethyltaurine. DMSP was found at 0.5-3 mmol/kg in all species with endosymbionts. Freshly isolated Symbiodinium from Fungia, P. damicornis, and P. compressa were also analyzed. DMSP and glycine betaine dominated in the first two; Porites endosymbionts had DMSP, proline betaine, and dimethyltaurine. In all specimens, glycerol and glucose were detected by high-performance liquid chromatography only at 0-1 mmol/kg wet mass. An enzymatic assay for glycerol plus glycerol 3-phosphate and dihydroxyacetone phosphate yielded 1-10 mmol/kg. Cassiopeia andromeda (upside-down jelly; Scyphozoan) and Aiptasia puchella (solitary anemone; Anthozoan) were also analyzed; both have endosymbiotic dinoflagellates. In both, glycine betaine, taurine, and DMSP were the dominant osmolytes. In summary, methylated osmolytes dominate in many Cnidaria; in those with algal symbionts, host and symbiont have similar methylated amino acids, as do congeners. However, little glycerol was present as an osmolyte and was probably metabolized before it could accumulate.