Abstract: A method for automated detection of landmarks from 3D medical image data using deep learning according to the present inventive concept, the method includes receiving a 3D volume medical image, generating a 2D intensity value projection image based on the 3D volume medical image, automatically detecting an initial anatomical landmark using a first convolutional neural network based on the 2D intensity value projection image, generating a 3D volume area of interest based on the initial anatomical landmark and automatically detecting a detailed anatomical landmark using a second convolutional neural network different from the first convolutional neural network based on the 3D volume area of interest.

Abstract: Systems, methods, and articles of manufacture are disclosed for learning models of movies, keywords, actors, and roles, and querying the same. In one embodiment, a recommendation application optimizes a model based on training data by initializing the mean and co-variance matrices of Gaussian distributions representing movies, keywords, and actors to random values, and then performing an optimization to minimize a margin loss function using symmetrical or asymmetrical measures of similarity between entities. Such training produces an optimized model with the Gaussian distributions representing movies, keywords, and actors, as well as shift vectors that change the means of movie Gaussian distributions and model archetypical roles. Subsequent to training, the same similarity measures used to train the model are used to query the model and obtain rankings of entities based on similarity to terms in the query, and a representation of the rankings may be displayed via, e.g., a display device.

Abstract: An article is described herein that includes: a glass-based substrate comprising opposing major surfaces; a crack mitigating composite over one of the major surfaces, the composite comprising an inorganic element and a polymeric element; and a hard film disposed on the crack mitigating composite comprising an elastic modulus greater than or equal to the elastic modulus of the glass-based substrate. The crack mitigating composite is characterized by an elastic modulus of greater than 30 GPa. Further, the hard film comprises at least one of a metal-containing oxide, a metal-containing oxynitride, a metal-containing nitride, a metal-containing carbide, a silicon-containing polymer, a carbon, a semiconductor, and combinations thereof.

Abstract: A method for automatically detecting a landmark in three-dimensional (3D) dental scan data includes projecting 3D scan data to generate a two-dimensional (2D) depth image, determining full arch data obtained by scanning all teeth of a patient and partial arch data obtained by scanning only a part of teeth of the patient by applying the 2D depth image to a convolutional neural network model, detecting a 2D landmark in the 2D depth image using a fully-connected convolutional neural network model and back-projecting the 2D landmark onto the 3D scan data to detect a 3D landmark of the 3D scan data.

Abstract: A method of separating RNA from a sample, comprising providing: a sample comprising RNA, a binding solution comprising an oligoethylene glycol and a salt, and a solid support having a hydrophilic surface. The method further comprises contacting the sample with the binding solution and solid support, under conditions that allow binding of the RNA in the sample to the surface of the solid support, thereby providing a solid support with bound RNA in contact with residual solution; and separating the solid support with bound RNA from the residual solution. During the binding of the RNA to the surface of the solid support, oligoethylene glycol is present at a concentration of at least about 35% v/v and the salt is present at a concentration of between about 1 M and about 2 M. Methods for producing purified RNA are also provided, comprising in vitro transcription to produce an RNA molecule with a first magnetic bead and purification with a second magnetic bead. Also provided are kits and uses.

Abstract: An automated registration method of 3D facial scan data and 3D volumetric medical image data using deep learning, includes extracting scan landmarks from the 3D facial scan data using a convolutional neural network, extracting volume landmarks from the 3D volumetric medical image data using the convolutional neural network and operating an initial registration of the 3D facial scan data and the 3D volumetric medical image data using the scan landmarks and the volume landmarks.

Abstract: An automated method for aligning a 3D dental library model to 3D oral scan data includes determining a valid tooth of the 3D oral scan data, extracting scan landmarks of the 3D oral scan data, loading a dental library model corresponding to the valid tooth of the 3D oral scan data, extracting library landmarks of the 3D dental library model, initial-aligning the 3D dental library model to the 3D oral scan data using the scan landmarks and the library landmarks and matching an individual tooth of the 3D dental library model and an individual tooth of the 3D oral scan data.

Abstract: A method for automated detection of landmarks from 3D medical image data using deep learning according to the present inventive concept, the method includes receiving a 3D volume medical image, generating a 2D intensity value projection image based on the 3D volume medical image, automatically detecting an initial anatomical landmark using a first convolutional neural network based on the 2D intensity value projection image, generating a 3D volume area of interest based on the initial anatomical landmark and automatically detecting a detailed anatomical landmark using a second convolutional neural network different from the first convolutional neural network based on the 3D volume area of interest.

Abstract: An automated method for aligning 3D (three-dimensional) dental data includes extracting landmark points of a CT (computerized tomography) data, extracting landmark points of scan data of a digital impression model, determining an up vector representing a direction of a patient's eyes and nose and identifying left and right of the landmark points of the scan data, extracting a teeth portion of the scan data, searching a source point of the scan data on a spline curve of the CT data to generate a candidate target point group and determining the candidate target point group having a smallest error with the landmark points of the CT data as a final candidate.

Abstract: Systems, methods, and articles of manufacture are disclosed for learning models of movies, keywords, actors, and roles, and querying the same. In one embodiment, a recommendation application optimizes a model based on training data by initializing the mean and co-variance matrices of Gaussian distributions representing movies, keywords, and actors to random values, and then performing an optimization to minimize a margin loss function using symmetrical or asymmetrical measures of similarity between entities. Such training produces an optimized model with the Gaussian distributions representing movies, keywords, and actors, as well as shift vectors that change the means of movie Gaussian distributions and model archetypical roles. Subsequent to training, the same similarity measures used to train the model are used to query the model and obtain rankings of entities based on similarity to terms in the query, and a representation of the rankings may be displayed via, e.g., a display device.

Abstract: A recoverable thrombosis filter that may be recoverable by a single recovery procedure. It can include a plurality of thrombosis filtering elements that are shaped in a predetermined manner and which are joined at one end and are deployed about a longitudinal axis to form a generally conical structure. The filtering elements can include shaped ends for engaging an inner lumen wall.

Abstract: A thrombus filter configured for placement within a blood vessel lumen defined by a blood vessel wall. The thrombus filter includes a body member, and a plurality of wires. Each wire has a joined end and a free end. The joined end of each wire is fixedly attached to the distal portion of the body member. Each wire radiates away from the body member along a generally helical path of expanding diameter. The shape of each wire may be generally described as a spiral, or a helix of expanding diameter. The wires radiate away from the body member to form a generally conical filtering portion which includes a plurality of open cells. The thrombus filter may be removed using minimally invasive methods when such removal becomes desirable. A method of removing the thrombus filter is disclosed.

Abstract: A recoverable thrombosis filter that can be implanted and securely positioned within a vein at a desired location, and can be recovered through an endovenous route even after formation of neointima hyperplasia, is disclosed.

Abstract: A recoverable thrombosis filter that can be implanted and securely positioned within a vein at a desired location, and can be recovered through an endovenous route even after formation of neointima hyperplasia, is disclosed.

Abstract: A thrombus filter configured for placement within a blood vessel lumen defined by a blood vessel wall. The thrombus filter includes a body member, and a plurality of wires. Each wire has a joined end and a free end. The joined end of each wire is fixedly attached to the distal portion of the body member. Each wire radiates away from the body member along a generally helical path of expanding diameter. The shape of each wire may be generally described as a spiral, or a helix of expanding diameter. The wires radiate away from the body member to form a generally conical filtering portion which includes a plurality of open cells. The thrombus filter may be removed using minimally invasive methods when such removal becomes desirable. A method of removing the thrombus filter is disclosed.