Catálogo de publicaciones - libros

Described here is an R&D project at the National Library of Medicine with the goal of creating systems to (a) provide the lay public images of the human anatomy, specifically high resolution color cryosections from NLM’s Visible Human Project and 3D images of anatomic structures created from these cryosections; (b) enhance text-based information services with relevant anatomic images. To accomplish these objectives, investigations into advanced techniques and technologies were conducted, including multi-tier system architectures, database design, design of suitable image viewers, image compression, and use of the Unified Medical Language System (UMLS), among others. This research has contributed to the design and development of AnatQuest, a system released for use by the lay public. It has also helped define the system architecture and essential functions required to link biomedical terms in documents to relevant anatomic structures in our database through UMLS concepts and relationships and to display these to the reader. In this chapter, we describe how our research has informed the overall goal to explore and implement new and visually compelling ways to bring anatomic images from the Visible Human dataset to the general public.

This chapter describes the emerging discipline of 3D Medical Informatics. While text-based informatics has a distinguished history and accepted fundamental linguistic principles, the use of 2D and 3D data in informatics has emerged relatively recently as computing capabilities have rapidly advanced, imaging and modeling standards have been established, and as high performance networking has made possible the sharing of the large and complex data that images, volumes, and models represent. This chapter outlines the early developments of this discipline, presents some examples of how image data is managed and presented, and suggests some of the leading research challenges in this young field.

Infectious disease informatics is an emerging field that studies data collection, sharing, modeling, and management issues in the domain of infectious diseases. This chapter provides an overview of this field with specific emphasis on the following two sets of topics: (a) the design and main system components of an infectious disease information infrastructure, and (b) spatio-temporal data analysis and modeling techniques used to identify possible disease outbreaks. Several case studies involving real-world applications and research prototypes are presented to illustrate the application context and relevant system design and data modeling issues.

Natural language processing is increasingly used to support biomedical applications that manipulate information rather than documents. Examples include automatic summarization, question answering, and literature-based scientific discovery. Semantic processing is a method of automatic language analysis that identifies concepts and relationships to represent document content. The identification of this information depends on structured knowledge, and in the biomedical domain, one such resource is the Unified Medical Language System. After providing some linguistic background, we discuss several semantic interpretation systems being developed in biomedicine. Finally, we briefly investigate two applications that exploit semantic information in MEDLINE citations; one focuses on automatic summarization and the other is directed at information extraction for molecular biology research.

Accessibility to a comprehensive variety of different types of structured patient data is critical to improvement in the health care process, yet most patient information is in the form of narrative text. Semantic methods are needed to interpret and map clinical information to a structured form so that the information will be accessible to other automated applications. This chapter focuses on semantic methods that map narrative patient information to a structured coded form.

Identification of relationships among different biological entities, e.g., genes, proteins, diseases, drugs and chemicals, etc, is an important problem for biological researchers. While such information can be extracted from different types of biological data (e.g., gene and protein sequences, protein structures), a significant source of such knowledge is the biological textual research literature which is increasingly being made available as large-scale public-domain electronic databases (e.g., the Medline database). Automated extraction of such relationships (e.g., gene A inhibits protein B) from textual data can significantly enhance biological research productivity by keeping researchers up-to-date with the state-of-the-art in their research domain, by helping them visualize biological pathways, and by generating likely new hypotheses concerning novel interactions some of which can be good candidates for further biological research and validation. In this chapter, we describe the computational problems and their solutions in such automated extraction of relationships, and present some recent advances made in this area.

Metabolic networks combine metabolism and regulation. These complex networks are difficult to understand and create due to the diverse types of information that need to be represented. This chapter describes a suite of interlinked tools for developing, displaying, and modeling metabolic networks. The metabolic network interactions database, MetNetDB, contains information on regulatory and metabolic interactions derived from a combination of web databases and input from biologists in their area of expertise. PathBinderA mines the biological “literaturome” by searching for new interactions or supporting evidence for existing interactions in metabolic networks. Sentences from abstracts are ranked in terms of the likelihood that an interaction is described and combined with evidence provided by other sentences. FCModeler, a publicly available software package, enables the biologist to visualize and model metabolic and regulatory network maps. FCModeler aids in the development and evaluation of hypotheses, and provides a modeling framework for assessing the large amounts of data captured by high-throughput gene expression experiments.

Automatically extracting gene-pathway relations from medical research texts gives researchers access to the latest findings in a structured format. Such relations must be precise to be useful. We present two case studies of approaches used to automatically extract gene-pathway relations from text. Each technique has performed at or near the 90 percent precision level making them good candidates to perform the extraction task. In addition, we present a visualization system that uses XML to interface with the extracted gene-pathway relations. The user-selected relations are automatically presented in a network display, inspired by the pathway maps created by gene researchers manually. Future research involves identification of equivalent relations expressed differently by authors and identification of relations that contradict each other along with the inquiry of how this information is useful to researchers.

The genomic data mine represents a fundamental shift from genetics to genomics, essentially from the study of one gene at a time to the study of entire genetic metabolic networks and whole genomes. Experimental laboratory data are deposited into large public repositories and a wealth of computational data mining algorithms and tools are applied to mine the data. The integration of different types of data in the genomic data mine will contribute towards an understanding of the systems biology of living organisms, contributing to improved diagnoses and individualized medicine. This chapter focuses on the genomic data mine consisting of text data, map data, sequence data, and expression data, and concludes with a case study of the Gene Expression Omnibus (GEO).

In this chapter, an introductory description of the exploration of genomic data is given. Rather than attempt an exhaustive overview of the types of genomic data and methods of analysis, the chapter focuses on one type of data, gene expression profiling by microarray technology, and one method of analysis, cluster analysis for discovering and sorting mixed populations. This type of data and method of analysis is very common in bioinformatics. It illustrates recurring problems and solutions. And a major portion of bioinformatics dealing with exploratory genomic data analysis can be viewed as a refinement and extension of this basic analysis.