Structures In 3D Space Research Articles

Deep Learning Representation using Autoencoder for 3D Shape Retrieval

We study the problem of how to build a deep learning representation for 3D shape. Deep learning has shown to be very effective in variety of visual applications, such as image classification and object detection. However, it has not been successfully applied to 3D shape recognition. This is because 3D shape has complex structure in 3D space and there are limited number of 3D shapes for feature learning. To address these problems, we project 3D shapes into 2D space and use autoencoder for feature learning on the 2D images. High accuracy 3D shape retrieval performance is obtained by aggregating the features learned on 2D images. In addition, we show the proposed deep learning feature is complementary to conventional local image descriptors. By combing the global deep learning representation and the local descriptor representation, our method can obtain the state-of-the-art performance on 3D shape retrieval benchmarks.

Leveraging protein quaternary structure to identify oncogenic driver mutations.

BackgroundIdentifying key “driver” mutations which are responsible for tumorigenesis is critical in the development of new oncology drugs. Due to multiple pharmacological successes in treating cancers that are caused by such driver mutations, a large body of methods have been developed to differentiate these mutations from the benign “passenger” mutations which occur in the tumor but do not further progress the disease. Under the hypothesis that driver mutations tend to cluster in key regions of the protein, the development of algorithms that identify these clusters has become a critical area of research.ResultsWe have developed a novel methodology, QuartPAC (Quaternary Protein Amino acid Clustering), that identifies non-random mutational clustering while utilizing the protein quaternary structure in 3D space. By integrating the spatial information in the Protein Data Bank (PDB) and the mutational data in the Catalogue of Somatic Mutations in Cancer (COSMIC), QuartPAC is able to identify clusters which are otherwise missed in a variety of proteins. The R package is available on Bioconductor at: http://bioconductor.jp/packages/3.1/bioc/html/QuartPAC.html.ConclusionQuartPAC provides a unique tool to identify mutational clustering while accounting for the complete folded protein quaternary structure.Electronic supplementary materialThe online version of this article (doi:10.1186/s12859-016-0963-3) contains supplementary material, which is available to authorized users.

The Constructive Approach to Teaching Anatomy through Clay Modeling: A Multi‐Tier Educational Value

The art of teaching anatomy has been traditionally performed via dissection. Most anatomy courses, ranging from fetal pig dissection in high school (HS) to human donor dissection in medical school, require a destructive process for discovering anatomical structures in 3D space. In contrast, clay modeling (CM) is a constructive approach to understanding spatial anatomical relationships. This qualitative study reports the experiences of two former medical students, a Syracuse HS teacher, anatomy faculty from two medical schools, and a third grader who participated in daylong workshops using the Anatomy in Clay Learning System. This haptic approach, used to model female perineal muscles from clay, emphasized touch and proprioception with continuous iterative feedback to ensure learning throughout the activity. While the instructors and students (now resident physicians) gained a greater appreciation of the region, a new world of anatomy was exposed to an unsuspecting, but curious, elementary school student. Participants with anatomy experience reported much improved familiarity with the pelvis, and in particular the attachments of the obturator internus muscle, even after having dissected the region previously. While the pedagogical question whether to implement CM before or after lab dissections arose, the experience revealed CM to be a highly valuable tool across multiple educational tiers.

Dealing with Aperiodic Protein Crystal Structures

Protein crystals can be aperiodic. They will diffract X-rays, and are therefore a crystal, but their diffraction is not periodic. This means that their diffraction pattern cannot be simply be indexed by a typical three-dimensional unit cell and space group. Aperiodic crystals include "quasi-crystals" as well as "modulated" crystals. In the latter case, the modulation can be positional or occupational and these modulations can be incommensurate with the normal periodic lattice [1]. An overview of aperiodic protein crystal diffraction will be presented with examples. The discussion will then focus on the characteristics of incommensurately modulated crystals followed by a more detailed discussion of how to solve these crystals. The following details of structure solution will be presented: (1) data collection perils; (2) specialized diffraction data processing in (3+1)D space using a q-vector [2]; (3) how to get an approximation of the structure in 3D space; (4) the assignment of the (3+1)D space group; and the ultimate (5) crystallographic refinement in superspace[3]. Future directions and needs will be discussed.

Occlusion-free feature exploration for volume visualization

Direct volume rendering is one of the most effective ways to visualize volume data sets, which employs intuitive 2D images to display internal structures in 3D space. However, opaque features always occlude other parts of the volume, and make some features of interest invisible in the final rendered images. Although a class of highly transparent transfer functions are capable of revealing all features at once, it is still laborious and time-consuming to specify appropriate transfer functions to reduce occlusions on less important structures and highlight features of interest, even for experts. Thus, the research on simpler volume visualization techniques which do not rely on complex transfer functions has been a hotspot in many practical applications. In this paper, an occlusion-free feature exploration approach that consists of modifying the traditional volume rendering integral is proposed, which can achieve better visibility of all internal features of interest with simple linear transfer functions. During the ray casting, a modulation parameter is derived to reduce the contributions of previous samples along the viewing ray, whenever the accumulated opacity value is close to overflow. In addition, several relevant functions are introduced to refine the modulation parameter and highlight the features of interest identified according to the attributes such as scalar, gradient module, occurrence and depth. Thereby, the proposed approach is capable of generating informative rendered images and enhancing the visual perception of features of interest without resorting to complex transfer functions.

3D Modeling and Visual Simulation of Plain Knitted Structures

Weft knitted fabrics are widely used in clothing textiles and technical applications, such as sweater and underwear, etc. The geometrical modeling of weft knitted structures is useful to predict fabrics properties and for creating physical models. How to realize the realistic computer simulation of plain knitted structures, much research is still needed. By measuring the loop structure size of fabrics produced by the circular machine and flat machine, a 3D geometrical model of plain loop with combination of empirical and theoretical was developed. The loop model has a total of 17 data points. Cubic NURBS curve was used to generate central path of yarn in plain loop. The circle was selected as the cross-section of yarn. In Visual C++ and OpenGL environment, then the 3D computer simulation of plain knitted fabric was realized. The simulation results distinctly show the intermeshed loops effects of knitted structures in 3D space.

Design of vascular networks: A mathematical model approach

In this paper, methods for the modeling of a realistic vascular tree in 3D space and for hemodynamic calculating throughout the simulated tree structure have been developed. Vascular trees are generated based on the power law relationship. Variations in branching asymmetry and segment length of the vasculature are precisely controlled by the designed gaussian distributions. The resolution limit of current imaging techniques for vessel detectability is simulated by designed pruning technique. On the basis of the generated diameters and lengths, the space locations of the vessel segments are calculated by optimizing the out-of-plane angles of two daughter branches. The generated vascular tree not only follows the power law relationship, but also maximizes the filling volume of the tree structure in 3D space. From the hemodynamic calculation in the simulated vasculature, the processes for which structural changes affect hemodynamic distributions are studied in detail. And also, the fractal nature and resistance of the vascular trees are quantified and compared. The developed method provides some insight into the design of the vascular trees in biology and may be used as a reference for the study of vascular diseases.

Processes in space

We introduce a geometric process algebra based on affine geometry, with the aim of describing the concurrent evolution of geometric structures in 3D space. We prove a relativity theorem stating that algebraic equations are invariant under rigid body transformations.

Swarm grammars: growing dynamic structures in 3D agent spaces

We present a new way of dynamically growing and breeding structures in 3D space through swarming agents. Different agent types and the way they evolve over time is specified by a swarm grammar similar to Lindenmayer systems. We expand common L-system string interpretation from a single turtle to a multitude of turtles which behave like a swarm. By describing swarm agents within the framework of formal grammars, we build a bridge from symbolic production systems (rewrite systems) to three-dimensional real-time construction procedures that are executed by reactive and interacting agents which move in simulated physical 3D spaces. We introduce constructor agents, their formal representation in swarm grammars and demonstrate by examples how (1) the swarm rules, (2) the agent parameters and (3) the environ ment can influence the actual construction and growth processes that are initiated and directed by the swarms. In order to facilitate exploration of a large variety of swarm grammars, we apply interactive evolutionary design methods to create swarm grammar sculptures and 3D structures.

Highly Oriented Thin‐Film Microdomain Patterns of Ultrahigh Molecular Weight Block Copolymers via Directional Solidification of a Solvent

Self-assembly is rapidly emerging as a simple and effective method for creating large-area functional periodic structures in various nanotechnologies. [1–3] In particular, block copolymers (BCPs) have been considered as promising self-organizing platforms because of their tunability in shape, size, and physical/chemical properties of the domains, with the capability of hosting many types of additives for desirable multifunctional composites. [4] Recently, ultrahigh molecular weight (number average molecular weight, Mn ∼10 6 gmol –1 ) BCPs possessing a domain size comparable to the wavelength of visible light have received increasing attention as self-assembled photonic materials. [5] 1D, 2D, and 3D photonic crystals for visible frequencies have been successfully demonstrated using lamellar, cylindrical, and double gyroid morphologies of polystyrene-block-polyisoprene (PS-b-PI) BCPs. [6–8] While most previous work on these photonic BCPs has been limited to relatively thick samples with only modest control over microdomain order, numerous engineering applications such as thinfilm waveguides, reflectors, and microcavities for lasing can potentially be achieved from controlled thin-film microstructures of these materials, all of which depend critically on the overall order of the microdomains to meet desired functionalities. Here we demonstrate the excellent control of thin-film microdomain patterns of lamellar- and cylinder-forming ultrahigh molecular weight BCPs (ca. 20 times larger molecular weight than for previously studied block copolymers) over a large area via directional solidification (DS) of a solvent. The ordering behavior of these photonic BCPs via the DS process has been found to be dramatically different from that of directionally solidified conventional molecular weight BCPs. Because of the relatively large domain sizes of ultrahigh molecular weight BCPs, laser scanning confocal microscopy (LSCM) could be used to optically characterize the lamellar and cylindrical thin film structures in 3D space.

Classical QSAR Study on Chromene Derivatives as Lanosterol 14α -Demethylase Inhibitor: A Non Azole Antifungal Target

A lanosterol 14alpha-demethylase inhibitors of chromene series were subjected to classical quantitative structural activity relationship studies. Apart from the indicator variables encoding for different group contribution, there are various physico-chemical descriptors like steric, thermodynamic and electronic parameters, which were applied to explore the structural requirements for inhibition of enzyme. Multiple linear regression analysis shows that substituents on the appended alkylated ether and the carbon chain length, which should not be more than six at R3 and R1 positions in the parent nucleus, are essential to modulate the activity. Electronic parameters such as highest occupied molecular orbital energy and dipole dipole energy have been found to play an important contribution for biological activity. From the orientation or distribution of the total molecular structures in 3D space, it was assessed that Principal Moments of Inertia at X-axis is detrimental for fungal inhibitory activity. Thus validated models bring important structural insights to aid the design of potent lanosterol 14alpha-demethylase inhibitors prior to their synthesis.

Interactive multi-scale visualizations of tonal evolution in MuSA.RT Opus 2

The purpose of this article is to describe MuSA.RT Opus 2, an interactive system for tonal visualization of music at multiple scales, and to present examples of the types of musical features and attributes that can be abstracted and visualized by the system. MuSA.RT aims to create an environment by which musical performances can be mapped in real-time to a concrete and visual metaphor for tonal space, wherein we can see the establishment and evolution of the tonal context. In this environment, expert musicians will be able to see the tonal structures of what they play, initiated listeners will be able to visually follow the structures that they hear, and novices can learn to hear the structures that they see. MuSA.RT is both an interactive art installation that can convert musical performances to mathematically elegant graphics and a scientific tool for visualizing the inner workings of tonal induction and tracking algorithms. In this article we describe the mapping strategies for transforming a MIDI stream into tonal structures in 3D space, and our solution for overcoming the challenge of real-time concurrent processing of data streams; we will also give examples and present case studies of visual mappings of music by Pachelbel, Bach, and Barber.

Stereological estimation of tubular length.

Very efficient and unbiased principles exist for estimating the total three-dimensional (3D), two-dimensional and zero-dimensional amounts of arbitrary structure in 3D space. The total one-dimensional length of real structure, in the ordinary sense, is an abstraction from the point of view of integral geometry. All stereological estimators of 'tubular length' are thus approximations. In addition, they are riddled by biases due to several types of artificial edges and other practical problems. This paper discusses several of these and proposes practical solutions of minimal biases.

Evolution and coevolution of developmental programs

The developmental processes of single organisms, such as growth and structure formation, can be described by parallel rewrite systems in the form of Lindenmayer systems, which also allow one to generate geometrical structures in 3D space using turtle interpretation. We present examples of L-systems for growth programs of plant-like structures. Evolution-based programming techniques are applied to design L-systems by Genetic L-system Programming (GLP), demonstrating how developmental programs for plants, exhibiting specific morphogenetic properties can be interactively bred or automatically evolved. Finally, we demonstrate coevolutionary effects among plant populations consisting of different species, interacting with each other, competing for resources like sunlight and nutrients, and evolving successful reproduction strategies in their specific environments.

A digital brain atlas for surgical planning, model-driven segmentation, and teaching

We developed a three-dimensional (3D) digitized atlas of the human brain to visualize spatially complex structures. It was designed for use with magnetic resonance (MR) imaging data sets. Thus far, we have used this atlas for surgical planning, model-driven segmentation, and teaching. We used a combination of automated and supervised segmentation methods to define regions of interest based on neuroanatomical knowledge. We also used 3D surface rendering techniques to create a brain atlas that would allow us to visualize complex 3D brain structures. We further linked this Information to script files in order to preserve both spatial information and neuroanatomical knowledge. We present here the application of the atlas for visualization in surgical planning far model-driven segmentation and for the teaching of neuroanatomy. This digitized human brain has the potential to provide important reference information for the planning of surgical procedures. It can also serve as a powerful teaching tool, since spatial relationships among neuroanatomical structures can be more readily envisioned when the user is able to view and rotate the structures in 3D space. Moreover, each element of the brain atlas is associated with a name tag, displayed by a user controlled pointer. The atlas holds a major promise as a template for model-driven segmentation. Using this technique, many regions of interest can be characterized simultaneously on new brain images.