Dual-Force ISOMAP: A New Relevance Feedback Method for Medical Image Retrieval

With great potential for assisting radiological image interpretation and decision making, content-based image retrieval in the medical domain has become a hot topic in recent years. Many methods to enhance the performance of content-based medical image retrieval have been proposed, among which the relevance feedback (RF) scheme is one of the most promising. Given user feedback information, RF algorithms interactively learn a user’s preferences to bridge the “semantic gap” between low-level computerized visual features and high-level human semantic perception and thus improve retrieval performance. However, most existing RF algorithms perform in the original high-dimensional feature space and ignore the manifold structure of the low-level visual features of images. In this paper, we propose a new method, termed dual-force ISOMAP (DFISOMAP), for content-based medical image retrieval. Under the assumption that medical images lie on a low-dimensional manifold embedded in a high-dimensional ambient space, DFISOMAP operates in the following three stages. First, the geometric structure of positive examples in the learned low-dimensional embedding is preserved according to the isometric feature mapping (ISOMAP) criterion. To precisely model the geometric structure, a reconstruction error constraint is also added. Second, the average distance between positive and negative examples is maximized to separate them; this margin maximization acts as a force that pushes negative examples far away from positive examples. Finally, the similarity propagation technique is utilized to provide negative examples with another force that will pull them back into the negative sample set. We evaluate the proposed method on a subset of the IRMA medical image dataset with a RF-based medical image retrieval framework. Experimental results show that DFISOMAP outperforms popular approaches for content-based medical image retrieval in terms of accuracy and stability.     

Mapping from GenBank to MEDLINE

GenBank has been based largely on literature that provides nucleic acid sequences. To find additional literature that is relevant to a given sequence, a search of MEDLINE can prove helpful. This paper documents some of the similarities between GenBank and MEDLINE that facilitate retrieval of documents from MEDLINE. In particular, techniques and examples are presented which take GenBank information and lead to MEDLINE information that supplements the GenBank information.     

MetaboSearch: tool for mass-based metabolite identification using multiple databases

Searching metabolites against databases according to their masses is often the first step in metabolite identification for a mass spectrometry-based untargeted metabolomics study. Major metabolite databases include Human Metabolome DataBase (HMDB), Madison Metabolomics Consortium Database (MMCD), Metlin, and LIPID MAPS. Since each one of these databases covers only a fraction of the metabolome, integration of the search results from these databases is expected to yield a more comprehensive coverage. However, the manual combination of multiple search results is generally difficult when identification of hundreds of metabolites is desired. We have implemented a web-based software tool that enables simultaneous mass-based search against the four major databases, and the integration of the results. In addition, more complete chemical identifier information for the metabolites is retrieved by cross-referencing multiple databases. The search results are merged based on IUPAC International Chemical Identifier (InChI) keys. Besides a simple list of m/z values, the software can accept the ion annotation information as input for enhanced metabolite identification. The performance of the software is demonstrated on mass spectrometry data acquired in both positive and negative ionization modes. Compared with search results from individual databases, MetaboSearch provides better coverage of the metabolome and more complete chemical identifier information. Availability: The software tool is available at http://omics.georgetown.edu/MetaboSearch.html.     

The interaction of positive and negative sensory feedback loops in dynamic regulation of a motor pattern

In many rhythmic behaviors, phasic sensory feedback modifies the motor pattern. This modification is assumed to depend on feedback sign (positive vs. negative). While on a phenomenological level feedback sign is well defined, many sensory pathways also process antagonistic, and possibly contradictory, sensory information. We here model the locust flight pattern generator and proprioceptive feedback provided by the tegula wing receptor to test the functional significance of sensory pathways processing antagonistic information. We demonstrate that the tegula provides delayed positive feedback via interneuron 301, while all other pathways provide negative feedback. Contradictory to previous assumptions, the increase of wing beat frequency when the tegula is activated during flight is due to the positive feedback. By use of an abstract model we reveal that the regulation of motor pattern frequency by sensory feedback critically depends on the interaction of positive and negative feedback, and thus on the weighting of antagonistic pathways.     

Improving life sciences information retrieval using semantic web technology

The ability to retrieve relevant information is at the heart of every aspect of research and development in the life sciences industry. Information is often distributed across multiple systems and recorded in a way that makes it difficult to piece together the complete picture. Differences in data formats, naming schemes and network protocols amongst information sources, both public and private, must be overcome, and user interfaces not only need to be able to tap into these diverse information sources but must also assist users in filtering out extraneous information and highlighting the key relationships hidden within an aggregated set of information. The Semantic Web community has made great strides in proposing solutions to these problems, and many efforts are underway to apply Semantic Web techniques to the problem of information retrieval in the life sciences space. This article gives an overview of the principles underlying a Semantic Web-enabled information retrieval system: creating a unified abstraction for knowledge using the RDF semantic network model; designing semantic lenses that extract contextually relevant subsets of information; and assembling semantic lenses into powerful information displays. Furthermore, concrete examples of how these principles can be applied to life science problems including a scenario involving a drug discovery dashboard prototype called BioDash are provided.     

BioGuideSRS: querying multiple sources with a user-centric perspective

Biologists are frequently faced with the problem of integrating information from multiple heterogeneous sources with their own experimental data. Given the large number of public sources, it is difficult to choose which sources to integrate without assistance. When doing this manually, biologists differ in their preferences concerning the sources to be queried as well as the strategies, i.e. the querying process they follow for navigating through the sources. In response to these findings, we have developed BioGuide to assist scientists search for relevant data within external sources while taking their preferences and strategies into account. In this article, we present BioGuideSRS, a user-friendly system which automatically retrieves instances of data by using BioGuide on top of the sequence retrieval system (SRS). BioGuideSRS is an Applet that can be run from its web page on any system with Java 5.0. AVAILABILITY: http://www.bioguide-project.net.     

Carryover effects from natal habitat type upon competitive ability lead to trait divergence or source–sink dynamics

Local adaptation to rare habitats is difficult due to gene flow, but can occur if the habitat has higher productivity. Differences in offspring phenotypes have attracted little attention in this context. We model a scenario where the rarer habitat improves offspring's later competitive ability – a carryover effect that operates on top of local adaptation to one or the other habitat type. Assuming localised dispersal, so the offspring tend to settle in similar habitat to the natal type, the superior competitive ability of offspring remaining in the rarer habitat hampers immigration from the majority habitat. This initiates a positive feedback between local adaptation and trait divergence, which can thereafter be reinforced by coevolution with dispersal traits that match ecotype to habitat type. Rarity strengthens selection on dispersal traits and promotes linkage disequilibrium between locally adapted traits and ecotype‐habitat matching dispersal. We propose that carryover effects may initiate isolation by ecology.     

A knowledgebase system to enhance scientific discovery: Telemakus

BACKGROUND: With the rapid expansion of scientific research, the ability to effectively find or integrate new domain knowledge in the sciences is proving increasingly difficult. Efforts to improve and speed up scientific discovery are being explored on a number of fronts. However, much of this work is based on traditional search and retrieval approaches and the bibliographic citation presentation format remains unchanged. METHODS: Case study. RESULTS: The Telemakus KnowledgeBase System provides flexible new tools for creating knowledgebases to facilitate retrieval and review of scientific research reports. In formalizing the representation of the research methods and results of scientific reports, Telemakus offers a potential strategy to enhance the scientific discovery process. While other research has demonstrated that aggregating and analyzing research findings across domains augments knowledge discovery, the Telemakus system is unique in combining document surrogates with interactive concept maps of linked relationships across groups of research reports. CONCLUSION: Based on how scientists conduct research and read the literature, the Telemakus KnowledgeBase System brings together three innovations in analyzing, displaying and summarizing research reports across a domain: (1) research report schema, a document surrogate of extracted research methods and findings presented in a consistent and structured schema format which mimics the research process itself and provides a high-level surrogate to facilitate searching and rapid review of retrieved documents; (2) research findings, used to index the documents, allowing searchers to request, for example, research studies which have studied the relationship between neoplasms and vitamin E; and (3) visual exploration interface of linked relationships for interactive querying of research findings across the knowledgebase and graphical displays of what is known as well as, through gaps in the map, what is yet to be tested. The rationale and system architecture are described and plans for the future are discussed.     

A framework for biomedical figure segmentation towards image-based document retrieval

The figures included in many of the biomedical publications play an important role in understanding the biological experiments and facts described within. Recent studies have shown that it is possible to integrate the information that is extracted from figures in classical document classification and retrieval tasks in order to improve their accuracy. One important observation about the figures included in biomedical publications is that they are often composed of multiple subfigures or panels, each describing different methodologies or results. The use of these multimodal figures is a common practice in bioscience, as experimental results are graphically validated via multiple methodologies or procedures. Thus, for a better use of multimodal figures in document classification or retrieval tasks, as well as for providing the evidence source for derived assertions, it is important to automatically segment multimodal figures into subfigures and panels. This is a challenging task, however, as different panels can contain similar objects (i.e., barcharts and linecharts) with multiple layouts. Also, certain types of biomedical figures are text-heavy (e.g., DNA sequences and protein sequences images) and they differ from traditional images. As a result, classical image segmentation techniques based on low-level image features, such as edges or color, are not directly applicable to robustly partition multimodal figures into single modal panels.In this paper, we describe a robust solution for automatically identifying and segmenting unimodal panels from a multimodal figure. Our framework starts by robustly harvesting figure-caption pairs from biomedical articles. We base our approach on the observation that the document layout can be used to identify encoded figures and figure boundaries within PDF files. Taking into consideration the document layout allows us to correctly extract figures from the PDF document and associate their corresponding caption. We combine pixel-level representations of the extracted images with information gathered from their corresponding captions to estimate the number of panels in the figure. Thus, our approach simultaneously identifies the number of panels and the layout of figures.In order to evaluate the approach described here, we applied our system on documents containing protein-protein interactions (PPIs) and compared the results against a gold standard that was annotated by biologists. Experimental results showed that our automatic figure segmentation approach surpasses pure caption-based and image-based approaches, achieving a 96.64% accuracy. To allow for efficient retrieval of information, as well as to provide the basis for integration into document classification and retrieval systems among other, we further developed a web-based interface that lets users easily retrieve panels containing the terms specified in the user queries.     

浅析医药卫生科技查新工作中存在的问题及对策

以南京军区福州总医院科技查新站为资料来源,浅析医药卫生科技查新工作中存在的相关问题:如科研人员查新意识淡薄及对查新流程不了解、查新工作特殊的阶段性问题、委托书中存在的问题、查新资源支持系统不够完善、查新人与课题委托人沟通不足、手工检索被忽略等,并结合查新实践提出解决问题对策以及提高查新质量的建议。     

Robust Prediction of t‐Year Survival with Data from Multiple Studies

Summary Recently meta‐analysis has been widely utilized to combine information across multiple studies to evaluate a common effect. Integrating data from similar studies is particularly useful in genomic studies where the individual study sample sizes are not large relative to the number of parameters of interest. In this article, we are interested in developing robust prognostic rules for the prediction of t ‐year survival based on multiple studies. We propose to construct a composite score for prediction by fitting a stratified semiparametric transformation model that allows the studies to have related but not identical outcomes. To evaluate the accuracy of the resulting score, we provide point and interval estimators for the commonly used accuracy measures including the time‐specific receiver operating characteristic curves, and positive and negative predictive values. We apply the proposed procedures to develop prognostic rules for the 5‐year survival of breast cancer patients based on five breast cancer genomic studies.     

Size-dependent interactions inhibit coexistence in intraguild predation systems with life-history omnivory

Growth in body size during ontogeny often results in changes in diet, leading to life-history omnivory. In addition, growth is often dependent on food density. Using a physiologically structured population model, we investigated the effects of these two aspects of individual growth in a system consisting of two size-structured populations, an omnivorous top predator and an intermediate consumer. With a single shared resource for both populations, we found that life-history omnivory decreases the likelihood of coexistence between top predator and intermediate consumer in this intraguild predation (IGP) system. This result contrasts with previous unstructured models and stage-structured models without food-dependent development. Food-dependent development and size-dependent foraging abilities of the predator resulted in a positive feedback between foraging success on the shared resource at an early life stage and foraging success on the intermediate consumer later in life. By phenomenologically incorporating this feedback in an unstructured IGP model, we show that it also demotes coexistence in this simple setting, demonstrating the robustness of the negative effect of this feedback.     

Prefrontal neural activity when feedback is not relevant to adjust performance

It has been shown that the rostral cingulate zone (RCZ) in humans uses both positive and negative feedback to evaluate performance and to flexibly adjust behaviour. Less is known on how the feedback types are processed by the RCZ and other prefrontal brain areas, when feedback can only be used to evaluate performance, but cannot be used to adjust behaviour. The present fMRI study aimed at investigating feedback that can only be used to evaluate performance in a word-learning paradigm. One group of volunteers (N = 17) received informative, performance-dependent positive or negative feedback after each trial. Since new words had to be learnt in each trial, the feedback could not be used for task-specific adaptations. The other group (N = 17) always received non-informative feedback, providing neither information about performance nor about possible task-specific adaptations. Effects of the informational value of feedback were assessed between-subjects, comparing trials with positive and negative informative feedback to non-informative feedback. Effects of feedback valence were assessed by comparing neural activity to positive and negative feedback within the informative-feedback group. Our results show that several prefrontal regions, including the pre-SMA, the inferior frontal cortex and the insula were sensitive to both, the informational value and the valence aspect of the feedback with stronger activations to informative as compared to non-informative feedback and to informative negative compared to informative positive feedback. The only exception was RCZ which was sensitive to the informational value of the feedback, but not to feedback valence. The findings indicate that outcome information per se is sufficient to activate prefrontal brain regions, with the RCZ being the only prefrontal brain region which is equally sensitive to positive and negative feedback.     

Multisite phosphorylation and network dynamics of cyclin-dependent kinase signaling in the eukaryotic cell cycle

Multisite phosphorylation of regulatory proteins has been proposed to underlie ultrasensitive responses required to generate nontrivial dynamics in complex biological signaling networks. We used a random search strategy to analyze the role of multisite phosphorylation of key proteins regulating cyclin-dependent kinase (CDK) activity in a model of the eukaryotic cell cycle. We show that multisite phosphorylation of either CDK, CDC25, wee1, or CDK-activating kinase is sufficient to generate dynamical behaviors including bistability and limit cycles. Moreover, combining multiple feedback loops based on multisite phosphorylation do not destabilize the cell cycle network by inducing complex behavior, but rather increase the overall robustness of the network. In this model we find that bistability is the major dynamical behavior of the CDK signaling network, and that negative feedback converts bistability into limit cycle behavior. We also compare the dynamical behavior of several simplified models of CDK regulation to the fully detailed model. In summary, our findings suggest that multisite phosphorylation of proteins is a critical biological mechanism in generating the essential dynamics and ensuring robust behavior of the cell cycle.     

A fast GPU-based hybrid algorithm for addition chains

A graphics processing unit (GPU) has been widely used to accelerate discrete optimization problems. In this paper, we introduce a novel hybrid parallel algorithm to generate a shortest addition chain for a positive integer e. The main idea of the proposed algorithm is to divide the search tree into a sequence of three subtrees: top, middle, and bottom. The top subtree works using a branch and bound depth first strategy. The middle subtree works using a branch and bound breadth first strategy, while the bottom subtree works using a parallel (GPU) branch and bound depth first strategy. Our experimental results show that, compared to the fastest sequential algorithm for generating a shortest addition chain, we improve the generation by about 70% using one GPU (NVIDIA GeForce GTX 770). For generating all shortest addition chains, the percentage of the improvement is about 50%.     

Noise Propagation in Gene Regulation Networks Involving Interlinked Positive and Negative Feedback Loops

It is well known that noise is inevitable in gene regulatory networks due to the low-copy numbers of molecules and local environmental fluctuations. The prediction of noise effects is a key issue in ensuring reliable transmission of information. Interlinked positive and negative feedback loops are essential signal transduction motifs in biological networks. Positive feedback loops are generally believed to induce a switch-like behavior, whereas negative feedback loops are thought to suppress noise effects. Here, by using the signal sensitivity (susceptibility) and noise amplification to quantify noise propagation, we analyze an abstract model of the Myc/E2F/MiR-17-92 network that is composed of a coupling between the E2F/Myc positive feedback loop and the E2F/Myc/miR-17-92 negative feedback loop. The role of the feedback loop on noise effects is found to depend on the dynamic properties of the system. When the system is in monostability or bistability with high protein concentrations, noise is consistently suppressed. However, the negative feedback loop reduces this suppression ability (or improves the noise propagation) and enhances signal sensitivity. In the case of excitability, bistability, or monostability, noise is enhanced at low protein concentrations. The negative feedback loop reduces this noise enhancement as well as the signal sensitivity. In all cases, the positive feedback loop acts contrary to the negative feedback loop. We also found that increasing the time scale of the protein module or decreasing the noise autocorrelation time can enhance noise suppression; however, the systems sensitivity remains unchanged. Taken together, our results suggest that the negative/positive feedback mechanisms in coupled feedback loop dynamically buffer noise effects rather than only suppressing or amplifying the noise.