Terminology proposal takes on Tower of Babel
Informatics may help radiologists comprehend MI's esoteric vocabulary
Charles Bankhead
Diag Imag, September 2003
http://www.diagnosticimaging.com/molecularimagingoutlook/2003sep/05.shtml
http://www.mi-central.org/
http://www.mi-central.org/papers/main.html
A Columbia University linguist has designed an informatics program that
helps radiologists understand how molecular imaging works. The model and the
standardized molecular imaging terminology associated with it are intended
to aid communications between molecular imaging researchers and geneticists,
according to Paola Karina Tulipano, a Ph.D. candidate in medical
informatics, who has spearheaded the effort. Tulipano and her colleagues are
formulating a terminology linked to genetic ontological concepts that will
help radiologists learn the language of molecular imaging and bring the
fields of imaging and genetics together.
The Human Genome Project has fueled an explosion of genetics research.
Within the past three years, a similar explosion has occurred in molecular
imaging, in particular imaging of gene expression, function, and products.
Although cross-pollination between the two fields would seem natural, the
lack of standardized terminology is an impediment.
"It would be great to start to develop and apply to molecular imaging
some
of the methods that have already been applied in bioinformatics," Tulipano
said. "We can use that information and also apply it to knowledge-mining
applications, so that computers can start to understand the information we
are looking for and use it to generate knowledge."
Tulipano, along with Dr. William S. Millar, a Columbia-Presbyterian Medical
Center radiologist, and Dr. James J. Cimino, a medical informatics
specialist, has proposed a concept-oriented terminology for the molecular
imaging field borrowed from the Gene Ontology Consortium. This organization
of molecular biology and bioinformatics experts develops structured
terminologies to permit annotation of genes, gene products, and gene
functions in bioinformatics databases.
Different laboratories discovered the same gene and gave it different
names," Tulipano said. "In an effort to integrate all of the information
generated by these different laboratories, bioinformaticians found they were
having a tough time deciphering the data. The Gene Ontology Consortium came
about from experts saying that we need a standard for all of this
information."
The ongoing efforts to develop terminology standards for molecular imaging
are analogous to the evolution of the DICOM electronic data communications
standard in radiology.
"DICOM has become the standard for information transfer, so people in
different regions and areas can look at images and talk the same language,"
said Millar, an assistant professor of radiology. "It seems like the perfect
time to get some sort of structuring of terminology in the molecular imaging
field started. Information needs to be structured so that we can all
understand it."
Tulipano presented her group's terminology proposal in January at the
Pacific Symposium on Biocomputing. The proposal is described in detail at a
Stanford University Medical Informatics Web site:
http://www.smi.stanford.edu/projects/helix/psb03/tulipano.pdf
The proposal evolved from a review of molecular imaging literature: Concepts
and terms were extracted from 30 different articles, ranging from broad
overviews to reports on specific applications. The investigators also
retrieved a published molecular imaging glossary to provide a starting point
for the collection of terms (Acad Radiol 2001;8:409-420).
The terms in the glossary were organized into a Frame-based representation
model with a hierarchical structure similar to that of the Medical Entities
Dictionary (J Am Med Inform Assoc 1994;1:35-50). Each concept in the
terminology was assigned a unique identifier, name, and named attributes.
In the hierarchical structure, the term molecular imaging entity emerged as
the top-level node concept with four direct descendants: imageable probe,
imageable target, amplification technique, and imaging instrument (see
figure). In turn, each of the four descendants from the top-level node
concept had direct descendants of its own.
A structured, or controlled, terminology has a variety of potential uses.
Radiologists and other physicians who want to learn more about molecular
imaging can use the terminology to retrieve research articles and other
information about the field. As the terminology evolves, information can be
organized to facilitate image retrieval and clinical documents with
associated images. The terminology might also be applied in ways that allow
tracking and structuring of image files according to use (e.g., clinical
versus teaching versus research). Medical libraries can apply the
terminology to the organization of information and resources related to
molecular imaging.
"Much of this is still theoretical, but I can envision a PACS in which
each
image stored in the system would have a field associated with it," Millar
said. "The field information could then be used to permit image searches,
so
information could be pulled from the system. PACS manufacturers might want
to consider this in the development of systems, so that information can be
tagged to images for later use."
The initial terminology represents a starting point. Tulipano hopes to get
input from experts in molecular imaging to develop a terminology that is
comprehensive and applicable to rapidly evolving fields of medicine and
technology.
--------
Linking molecular imaging terminology to the gene ontology (GO)
Tulipano PK, Millar WS, Cimino JJ.
Department of Medical Informatics, Columbia University, New York, NY 10032,
USA. tulipano@dmi.columbia.edu
Pac Symp Biocomput. 2003;:613-23.
http://www.smi.stanford.edu/projects/helix/psb03/tulipano.pdf
The rapidly developing domain of molecular imaging represents the merging of
current advances in the fields of molecular biology and imaging research.
Despite this merger, an information gap continues to exist between the
scientists who discover new gene products and the imaging scientists who can
exploit this information. The Gene Ontology (GO) Consortium seeks to provide
a set of structured terminologies for the conceptual annotation of gene
product function, process and location in databases. However, no such
structured set of concept-oriented terminology exists for the molecular
imaging domain. Since the purpose of GO is to capture the information about
the role of gene products, we propose that the mapping of GO's established
ontological concepts to a molecular imaging terminology will provide the
necessary bridge to fill the information gap between the two fields. We have
extracted terms and definitions from an already published molecular imaging
glossary as well as molecular imaging research articles, and developed
molecular imaging concepts. We then mapped our molecular imaging concepts to
the existing gene ontology concepts as a method to comprehensively represent
molecular imaging.
------------------------------
SOFG - Standards and Ontologies for Functional Genomics
http://www.sofg.org/
This web site provides a community resource for the development of
ontologies and other controlled vocabularies relevant to biology and
biological experimentation. It was established as a result of the first
conference on Standards and Ontologies for Functional Genomics (SOFG), which
was held at the Wellcome Trust Genome Campus, Hinxton, UK, 17-20 November
2002.
The goal of the SOFG web site is the same as that of the conference: "to
bring together biologists, bioinformaticians, and computer scientists who
are developing and using standards and ontologies with an emphasis on
describing high-throughput functional genomics experiments".
Ontologies for human and mouse anatomy:
This web site is part of a community effort to integrate ontologies for
human and mouse anatomy. Numerous ontologies for human and mouse anatomy
exist or are being developed. Each has its own purpose. For the biologist
who wants to annotate data with anatomical names this variety is confusing.
As a first step towards improving the situation, we will list the main
resources available to the biomedical and bioinformatics communities.
Several major groups have already joined this effort, but the list presented
on these pages is only a beginning.
--------------------------
Conference Report: Standards and ontologies for functional genomics: Towards
unified ontologies for biology and biomedicine
Wellcome Trust Genome Campus, Hinxton, Cambridge, UK, 17aEUR"20 November
2002
Midori A. Harris and Helen Parkinson*
European Bioinformatics Institute, EMBL Outstation, Wellcome Trust Genome
Campus, Hinxton, Cambridge CB10 1SD, UK
Comparative and Functional Genomics
Comp Funct Genom 2003; 4: 116-120.
Published online in Wiley InterScience (http://www.interscience.wiley.com).
DOI:
10.1002/cfg.249
In recent years, whole genome analysis has become routine and systems
biology and modelling of whole cells are becoming more common. Advances in
experimental technology now permit expression analysis for tens of thousands
of genes at a time, generating vast amounts of biological data, and the
application of high-throughput technologies to
proteomics makes the burden even heavier. Databases and tools have become
available to help biologists manage and interpret these large volumes of
data, but databases alone are not sufficient to allow researchers to
integrate large amounts, and different kinds, of information. The problem is
that any given biological phenomenon can be described in many different
ways; an example is the use of free text annotation in databases such as
GenBank, which has resulted in inconsistent data representation. One
promising solution is to provide consistent annotation within databases by
means of standard formats and ontologies. Standards development depends on
the availability of suitable ontologies; both the MIAME standard and
emerging proteomics standards recommend the use of ontologies wherever
possible. Where such ontologies do not exist, communities are starting to
build their own to support standards.
Although the exact meaning of the word ontology is contentious, T.Gruber's
definition of an ontology as a ~specification of a conceptualization"
(http://www-ksl.stanford.edu/kst/what-isan-ontology.html)
is widely used.
Ontologies are, at a minimum, sets of terms used in a specific domain,
definitions for those terms and defined relationships between the terms;
they can range from simple controlled vocabularies to structurally complex
representations employing description logics. In biology, the use of
ontologies allows database annotation to be standardized, and makes
sophisticated
queries possible for humans and computers.
The SOFG conference brought together about 120 biological domain experts,
computer scientists and those with interests in related fields, such as
natural language processing (NLP), to address the issues of developing,
implementing and ultimately unifying ontologies.
[Full text version of this paper is available at:
http://www.hgmp.mrc.ac.uk/CCP11/tbrpapers/CurrentPapers/sofg.pdf
]
-----------------------
SOFG Conference Report
Andrew Jones
http://www.dcs.gla.ac.uk/~jonesa/ConfReport/ConfReport.html
The conference was intended to attract researchers presenting work related
to the development of ontologies for describing biology and biomedicine. The
conference was attended by approximately 140 delegates, with a poster
sessions including 40 posters.
There were seven oral sessions:
* Introduction to Ontologies
* GO and GOBO
* Ontologies for Model Organisms
* Tools for Building Ontologies
* Ontologies for Medicine and Pathology
* Implementation and Use of Ontologies
* Ontologies for Chemistry, Toxicology and other Domains
Copies of the presentations can be obtained from the SOFG website at the
EBI. Keynote speeches were given by Ken Buetow, Lincoln Stein, Winston Hide
and Peter Karp. Ken Buetow is director of the NCICB (the National Cancer
Institute, Centre for Bioinformatics.
Peter Karp described the development of a collection of bacterial pathway
databases known as BioCyc. There are two types of databases included in
BioCyc, databases derived from the literature and databases of pathways
derived computationally from genome analysis. The computationally derived
databases are created using software known as Pathologic, that takes as
input an annotated genome sequence and outputs a new pathway database. There
are currently 13 derived databases for different bacterial species. There
are two databases created manually EcoCyc: for E.coli pathways have been
inferred from published experimental data, and MetaCyc: which includes
metabolic data derived from literature for 150 different organisms.
Lincoln Stein gave an entertaining presentation discussing the features of
human nature to classify, using the universal popularity of the Japanese
game Pokemon as his main example.
Winston Hide is from SANBI (South Africa National Bioinformatics Institute,
his presentation detailed the work at their institute to develop ontologies
for describing gene expression experiments, such as SAGE and EST libraries.
The research uses ontologies for anatomical systems, cell types,
developmental stages and pathology to map between the annotation that is
given to EST or SAGE data sets, to allow queries to search for specific
types of experiment. The talk also described the advantages of an approach
using open standards and open source software, such as improved access to
resources and sharing of data.
--------------------------------------
Proceedings of the Workshop on Health Informatics Research and Development
Hosted by:
Medical Research Council
Co-sponsored by:
Biotechnology and Biological Sciences Research Council
Engineering and Physical Sciences Research Council
Department of Health
Department of Trade and Industry
18 July 2002
Royal College of Obstetricians and Gynaecologists
London
http://www.mrc.ac.uk/pdf-health_informatics_r_and_d_workshop.pdf
http://www.mrc.ac.uk/index/strategy-strategy/strategy-science_strategy/strategy-strategic_implementation/strategy-e_science_highlight_notices/strategy-health_informatics_workshop.htm
http://www.mrc.ac.uk/index/strategy-strategy/strategy-science_strategy/strat
egy-strategy_implementation/strategy-government_spending_review_initiatives/
strategy-e_science_highlight_notices/strategy-health_informatics_workshop_pr
esentations.htm
Health Informatics workshop
Royal College of Obstetricians and Gynaecologists, London, 18 July 2002
Provision of the best possible healthcare is critically dependent on
capturing the most complete data possible for accurate risk assessments (eg
genetic counselling, epidemiology, and analysis of disease patterns).
Preventative health measures resulting from this would both benefit the
individual patient and be more cost effective.
Health Informatics encompasses the organisational, professional, scientific
and technical issues involved in the use of information systems to support
patient centred healthcare, and is closely linked with healthcare
informatics. The diversity and complexity of clinical data (a combination of
text, codes, speech and images) makes the capture, storage and analysis of
this information one of the most exciting challenges of the new millennium.
However, maximal exploitation of the current explosion of information and
knowledge requires the integration of multiple data sources and
interoperability of different systems. Grid technology can provide the
necessary infrastructure.
By its very nature, health informatics is multidisciplinary and cuts across
many scientific boundaries. It involves industry, academia and the health
services, and often raises complex IPR issues. Moreover, it is highly
dependent on healthcare and other infrastructure, and is resource intensive.
Different aspects of health informatics fall within the remit of different
funding agencies, and there is currently no cohesive approach to funding.
Due in part to the complexity and diversity of the field, it is neither
possible nor desirable to try to address all the outstanding questions.
Rather, there is a need to identify and prioritise - both for the short and
longer term - those areas that will best advance the field.
The aim of this workshop, which was co-sponsored by the BBSRC, EPSRC, the
Core e-Science Programme, the Department of Health and the Department of
Trade and Industry, was to bring together experts from different areas to
discuss the key issues in health informatics; to learn from experiences,
identify current UK strengths and weaknesses, explore ways of addressing the
gaps and discuss how to take the field and the technical challenges forward.
In this way, both the scientific community and the funding agencies would
hopefully be able to develop a more cohesive research and funding strategy.
------------------------------
Nature - Imaging in Cell Biology
http://www.nature.com/focus/cellbioimaging/
Microscopy has been a key tool for cell biologists from the outset aEUR"
indeed,cell biology was literally born with microscopy. This year marks the
300th
anniversary of the death of Robert Hooke, whose seminal observations under
the microscope (Micrographia, 1665) led to the initial coining of the term
'cell' (referring to a tiny bare room, similar to a monk's cell). Since
then, microscopy has revolutionized our understanding of how cells live and
die. New subcellular compartments have been discovered thanks to improving
microscopy techniques, and progress in cell biology still relies in great
part on advances in imaging techniques.
This past decade has seen a 'rainbow' revolution in microscopy. Fluorescent
proteins, such as green fluorescent protein from the jellyfish Aequorea
victoria, have been used to visualize biological processes as they happen in
living cells and whole organisms. Together with improved fluorescence
microscopy and time-lapse microscopy, these impressive techniques have
provided insights into the dynamics of proteins and the biological processes
that they regulate. Importantly, imaging techniques are now becoming
available not only to specialist biophysicists but also to cell biologists
aEUR"
a merger that is reflected by joint conferences and collaborations that
bring these communities together. Now is therefore the perfect time to focus
on this central tool in cell biology.
For this reason, Nature Cell Biology and Nature Reviews Molecular Cell
Biology are pleased to present this supplement on imaging in cell biology.
These journals have sole responsibility for the choice and content of the
supplement and, should you wish to cite any of these articles, please refer
to the citation information at the end of each article, above the reference
list. The supplement consists of a series of specially commissioned
articles, which were selected on the basis of feedback from the research
community. The topics covered in the six Review articles span both the
wavelength and the resolution scale, from magnetic resonance imaging to
electron microscopy, and from single-molecule to whole-organism imaging. To
conclude this supplement, Roger Y. Tsien speculates on the future of imaging
and the new challenges that lie ahead. We hope that the content of these
articles provides not only an essential guide to the latest techniques,
their advantages and limitations, but also highlights the diverse
cell-biological applications of imaging techniques.
We are pleased to acknowledge the financial support of Carl Zeiss in the
production of this supplement. Thanks to this support, supplement articles
are available free online for six months, where they can be found together
with a Focus on imaging in cell biology (see
http://www.nature.com/focus/cellbioimaging).
In addition to the supplement
articles, this Focus site contains key imaging articles from our past issues
and from other Nature Publishing Group journals, as well as movies, images
and recommended links. The latter includes a link to the ongoing Nature Cell
Biology and Nature Reviews Molecular Cell Biology 'Cell of the Month'
competition, and some of the winning images have been used in this
supplement to emphasize that cell-biological images can be not only
informative, but also visually stunning.
ALISON SCHULDT, Associate Editor, Nature Cell Biology
RACHEL SMALLRIDGE, Senior Editor, Nature Reviews Molecular Cell Biology
-----------------
European Union funding aids innovative medical IT
Paula Gould
http://www.medicaltech.org/
The maxim "united we stand, divided we fall" is holding true for
a group of
medical IT researchers. Members of a Europe-wide consortium who joined
together in a single bid for European Union R&D funding are now seeing their
theoretical concepts realized as demonstrable clinical systems.
The consortium, known as EUTIST-M, consists of 39 European companies,
universities, and technology centers. It is currently funded under the
European Commission's so-called fifth framework program, a scheme designed
to promote excellence in European R&D by facilitating cross-border
collaboration and a pooling of intellectual resources.
EUTIST-M started three years ago as a collection of six medical IT projects,
according to Dr. Ignacio Blanquer, an associate professor at the Polytechnic
University of Valencia in Spain and a key player in the consortium. The
cluster gradually expanded to cover 11 separate R&D ventures, with
applications spanning radiology, surgery, orthopedics, oncology,
dermatology, audiology, and intensive care medicine. Consortium members meet
every six months for a roundtable discussion, and share a single Web site
(http://www.medicaltech.org) to publicize
their work.
The group approach has many advantages for researchers making the step from
idea to prototype product, Blanquer said. The twice-yearly meetings provide
a valuable forum for different project teams to exchange useful information,
while the overall costs of setting up clinical demonstrations and marketing
successful systems are reduced.
"Some of the partners didn't feel comfortable at first with the idea of
widespread dissemination of what we are doing, but this publicity is an
important aspect," he said.
Three of the 11 prototypes developed involve the management or manipulation
of medical images. Project VISU uses 3D CT image data and a haptic device to
help predict the outcome of craniofacial surgery. Virtual surgery can be
performed in about 15 minutes on a standard PC, allowing trainees to
practice new techniques and experienced consultants to experiment with
alternative approaches.
Project DISMEDI describes a low-cost tool for 3D segmentation and 3D or
multiplanar projection, navigation, and measurement during radiological
diagnosis. Project CREAM, on the other hand, outlines a system for
digitizing, DICOM-encoding, and storing imaging movies generated on analog
angiography and echocardiography equipment.
Researchers involved with the consortium must now make plans to stand on
their own two feet. Funding for EUTIST-M will come to an end in six months,
and changes to the EU's funding program mean that the consortia will no
longer be eligible for financial support.
"We are in the process of consolidating our database of contacts,"
Blanquer
said. "We will make visits to different contacts and arrange small
demonstrations of the prototypes specifically for each one."
-------------
EUTIST-M: Medical IT at Europe
http://www.medicaltech.org/
Research and Development activities on the Medical Sector are being carried
out by a consortium of 39 European Companies, Universities and Technology
Centres for improving the productiveness of medical information
technologies. The application areas are Radiology, Ear Fitting,
Orthopaedics, Oncology, Intensive Care Units, Surgery, Dermatology.
WELCOME to the EU Clusters Portal
In the EC Fifth Framework Programme, several clusters of projects have been
set up to maximise the efficiency of the coordination actions and to exploit
the synergies among the projects. The following four clusters are focused on
Agents and Middleware (EUTIST-AMI), Machine Vision (EUTIST-IMV), Application
Service Provision (ASP-BP), and Information Technology for Medicine
(EUTIST-M)
EUTIST-AMI, funded by European Union, is constituted by 17 application
projects that are testing the benefits and the potential of Agent and
Middleware technologies applied in real industrial environments.
ASP-BP is an EC-funded European framework which groups and manages 6
projects focused on ASP technology-based applications. It groups 30 partners
from 9 different European countries, with the aim of demonstrating the
business benefits of ASP Technology in a variety of industrial sectors
Integrated Machine Vision (IMV) systems bring major benefits to industry by
enabling better quality control, measurement and monitoring. In the
EC-supported initiative EUTIST-IMV over 80 industrial and academic partners
are developing new innovative solutions for real-life industrial problems.
EUTIST-M is a consortium of currently 11 medical projects funded by the IST
Programme of the European Commission that aim at fostering medical
information technologies by the development of five trials and one best
practice.