INITIAL RESULTS ON KNOWLEDGE DISCOVERY AND DECISION
SUPPORT FOR INTRACRANIAL ANEURYSMS
Christoph M. Friedrich, Martin Hofmann-Apitius
Fraunhofer Institute for Algorithms and Scientific Computing (SCAI), Department of Bioinformatics
Schloss Birlinghoven 53754 Sankt Augustin Germany
Robert Dunlop
Infermed Ltd, 25 Bedford Square, London, UK
Ioannis Chronakis
Department of Engineering, Oxford University, UK, Department of Academic Oncology, UCL, UK
Miriam C. J. M. Sturkenboom, Roelof Risselada
Erasmus MC, Medical Informatics, 3000 CA Rotterdam, Netherlands
Baldo Oliva, Ferran Sanz
Research Unit on Biomedical Informatics (GRIB) IMIM/UPF, C/Dr. Aiguader, 88, 08003 - Barcelona, Spain
Keywords:
Knowledge Discovery, Decision Support, Intracranial Aneurysm.
Abstract:
Intracranial Aneurysms are bulbous expansions of the intracranial vessels, that may rupture and lead to sub-
arachnoid haemorrhage, which can result in severe disability or death of the affected person. The prediction of
the individual rupture risk of a patient based on information from images, haemodynamic simulations, clinical
parameters and genetic markers is one of the aims of the European Integrated Project @neurIST. The pre-
dicted rupture risk is meant to support decision making on clinical treatment. We will present initial results on
Knowledge Discovery through a combination of text-mining, data integration from public bioinformatics data
sources, and database mining. Additionally, we provide first results for decision support through knowledge
based clinical guidelines and Bayesian networks.
1 INTRODUCTION
The advent of improved medical imaging facilities
and their routine use in clinical practice increases
the number of accidentally detected asymptomatic In-
tracranial Aneurysms (IA). Intracranial Aneurysms
are bulbous expansions of the intracranial vessels, that
may rupture and lead to intracranial bleeding (sub-
arachnoid haemorrhage), which can result in severe
disability or death of the affected person. In (Rinkel
et al., 1998) the prevalence of the disease for adults
without risk factors for subarachnoid haemorrhage is
reported with approximately 2 % and the annual risk
rate for a rupture with 0.7 %. This relatively high
prevalence with low incidence of the dangerous event
leads to controversial discussions on treatment deci-
sions. In general there are three treatment options:
1. do not treat the asymptomatic aneurysm with low
risk
2. conduct neurosurgical clipping
3. deploy a platinum coil via endovascular interven-
tion
One of the targets of the European Integrated
Project @neurIST
1
is to support decision making on
IA treatment options by building a distributed envi-
ronment for healthcare. This environment will al-
low access to patient related information from images,
1
http://www.aneurist.org
265
M. Friedrich C., Hofmann-Apitius M., Dunlop R., Chronakis I., C. J. M. Sturkenboom M., Risselada R., Oliva B. and Sanz F. (2008).
INITIAL RESULTS ON KNOWLEDGE DISCOVERY AND DECISION SUPPORT FOR INTRACRANIAL ANEURYSMS.
In Proceedings of the First International Conference on Health Informatics, pages 265-272
Copyright
c
SciTePress
haemodynamic simulations, clinical parameters, ge-
netic markers and epidemiological data.
Additionally, a set of application suites will be de-
veloped, that are based on this infrastructure and di-
rectly support the goal of improving clinical decision
making. A draft architecture of the distributed sys-
tem is described in (Arbona et al., 2007). Considering
the target of this paper, two application suites are of
interest, @neuRisk, a decision support system based
on clinical guidelines and @neuLink, a research ori-
ented application targeted at linking genetics to dis-
ease. The @neuLink suite supports Knowledge Dis-
covery for the detection of genetic risk factors.
2 INITIAL RESULTS
In the following we will give examples and prelimi-
nary results of Decision Support and Knowledge Dis-
covery that have been developed during the first year
of the project.
2.1 Decision Support based on the
Proforma Language and the
REACT Application
In this section we will describe the work carried out
in the development of the first @neuRisk prototype.
This version of the prototype employs a mixed quan-
titative and qualitative approach to provide risk as-
sessment and decision support in the treatment of
cerebral aneurysms. The knowledge base for the
approach is derived from two trials: The Interna-
tional Study on Unruptured Intracranial Aneurysms
(ISUIA) (Wiebers et al., 2003) and International Sub-
arachnoid Aneurysm Trial (ISAT) (Molyneux et al.,
2005). For demonstration purposes, some additional
test data were also included to show how future
research results could be incorporated to the final
@neuRisk suite.
Qualitative decision support in @neuRisk has
been implemented using the PROforma method and
tools (Sutton and Fox, 2003) while quantitative deci-
sion support was implemented by adapting an existing
treatment planning application called REACT (Risk,
Events, Actions and their Consequences over Time)
(Glasspool et al., 2006).
PROforma is a well established technology, first
published in 1996 (Fox et al., 1996) and described in
detail in 2000 (Fox and Das, 2000; Sutton and Fox,
2003), is a well established clinical decision support
technology, that has been tested in a number of trials
with promising results (Fox et al., 2006; Hurt et al.,
2003; Patkar et al., 2006). There are two major imple-
mentations available, the Tallis implementation from
Cancer Research UK
2
and the Arezzo implementation
by InferMed Ltd in London
3
. REACT technology is
based on PROforma concepts and has been tested in
one trial in the area of genetic counselling with en-
couraging results (Glasspool et al., 2006).
In the qualitative part of the prototype, we used
the PROforma language (Sutton and Fox, 2003) to
model both the workflow involved in patient manage-
ment and a set of treatment decisions. The resulted
computer-executable guideline application is then en-
acted by the Tallis PROforma engine. It guides the
user through the workflow, provides a set of data cap-
ture services to collect data from the various disci-
plines (clinician, radiologist, geneticist, etc) and fi-
nally, it offers support for the treatment decisions. It
does this by offering a set of logical arguments (rules)
to the clinicians which either support or oppose each
of the available treatment actions. The system sug-
gests the most appropriate action but the final decision
is taken by the clinician (Fox et al., 2006).
In the quantitative part of the prototype, we used
an adapted version of the REACT tool (Glasspool
et al., 2006). This tool provides support for planning
the treatment of the patient based on the effect that
each treatment action has on the risk. The REACT
user interface is divided into 4 major parts:
1. The treatment plan,
2. The graph area,
3. The argumentation area and
4. The notification area.
The treatment plan area provides the clinician with
a set of available treatment options and a timeline
where she/he can schedule them, similar to a Gantt
chart. As the user adds or removes events from the
treatment plan, the graph area plots the expected con-
sequences in real time (in the @neuRisk prototype
these are life expectancy and years gained or lost if an
aneurysm were left untreated). This allows the user
to explore the space of available options with imme-
diate feedback of the various interactions and conse-
quences. Information directly relevant to each option
2
Tallis is an implementation of the PROforma engine
written in Java which was developed at Cancer Research
UK by the Advanced Computation Laboratory (http://
acl.icnet.uk/) that was led by Prof John Fox. The en-
gine is supported by a suite of tools including an engine,
an authoring tool, a tester tool and a web based enacting
application (http://www.cossac.org/tallis.html).
3
Arezzo is an implementation of the PROforma engine
created by InferMed (http://www.infermed.com) and it
has been used in a number of commercial products.
HEALTHINF 2008 - International Conference on Health Informatics
266
pert work, specialisation on one topic and possible se-
lection bias. As an alternative to this manual extrac-
tion, we consider text-mining (Jensen et al., 2006).
This helps to get an overview on genes possibly in-
volved in a disease and to find potential new genes
from publications. We implemented a Find Candi-
date Genes module in the @neuLink application suite.
This part of the application suite is based on two text-
mining systems, ProMiner (Hanisch et al., 2005) and
OSIRIS (Bonis et al., 2006). ProMiner finds enti-
ties (Genes/Proteins, Drugnames, Chromosomal Lo-
cations . . . ) and links them to unique database identi-
fiers, e.g. EntrezGene (Maglott et al., 2005). OSIRIS
finds and disambiguates mentions of genetic varia-
tions in text to dbSNP (Smigielski et al., 2000) iden-
tifiers with a query-expansion approach.
To support the focussed view of the user on rel-
evant information in the disease context, we devel-
oped a ranking mechanism based on Relative Entropy
(Kullback and Leibler, 1951), also known as Kull-
back/Leibler divergence. In this ranking mechanism,
we use the complete MedLine as reference corpus and
contrast it with the specific corpus derived from full-
text search.
Finding and disambiguating variation mentions in
text with the OSIRIS system, needs a high-quality
gene-mention machinery. We therefore combined our
text-mining tools and complemented them with a ma-
chine learning variation finding engine based on Con-
ditional Random Fields (CRF) (Lafferty et al., 2001).
The improved results of this approach have been de-
scribed in (Klinger et al., 2007).
One of the crucial questions of all Discovery
methods is their validation. For the finding and dis-
ambiguation of gene mentions in text, we have been
able to get an independent assessment of the perfor-
mance of our approach by participating in the BioCre-
aTive II assessment (Morgan and Hirschmann, 2007).
Our ProMiner system assessed as described in (Fluck
et al., 2007), has been ranked 3rd of 21 submissions.
In a second evaluation, we tested wether our sys-
tem, given the keyword search “intracranial AND
aneurysm*”, was able to detect the same related sus-
pectibility genes, that have been found by human
experts. The review on genetics (Krischek and In-
oue, 2006) mentions 18 associated genes in the con-
text of Intracranial Aneurysms. In our evaluation (as
of 2007-10-01) (Gattermayer, 2007), we find 16,548
documents in PubMed related to the keyword and 596
documents, that mention 316 different genes/proteins.
We find and could disambiguate all 18 genes in pub-
lications and rank them to the first 238 hits with 7
hits among the top 16 candidates. See figure 3 for a
screenshot of the interface. Among the high-ranked
false positives we find frequently used therapeutic
proteins like the plasminogen activator (PLAT), but
also new true positives like the JAG1 gene, that have
not been mentioned in the genetic reviews.
2.5 Generating Protein-Protein
Interaction Networks
We used “Protein ineractions and network analysis”
(PIANA) (Aragues et al., 2006) to combine data
from the Database of Interacting Proteins (DIP) (Sal-
winski et al., 2004), the MIPS database of interac-
tions (Pagel et al., 2005), the Molecular INTeractions
database (MINT) (Chatr-aryamontri et al., 2007), In-
tAct (Kerrien et al., 2007), the Biomolecular In-
teractions Database (BIND) (Alfarano et al., 2005),
the BioGrid (Stark et al., 2006) and Human Pro-
tein Reference Database (HPRD) (Peri et al., 2003)
and the human interactions from two recent high-
throughput experiments (Rual et al., 2005),(Stelzl
et al., 2005). We also provide the interactions
obtained from STRING (von Mering et al., 2005)
and methods of protein-protein interaction prediction
based on sequence/structure patterns (Espadaler et al.,
2005), (Cockell et al., 2007). The integration of many
different sources of interactions into a single repos-
itory allowed us to work with an extensive set of
363,571 interactions between 42,040 different protein
sequences.
PIANA represents protein interaction data as a
network where the nodes are proteins and the edges
interactions between them. In such a network, a set
of proteins linked to protein p
j
(i.e. physically inter-
acting with p
j
) is named “partners of p
j
”. PIANA
builds the network by retrieving partners for a initial
set of seed proteins (i.e. the relevant proteins, here re-
ferred as “seed proteins”) that were obtained from the
Find Candidate Genes module in section 2.4. A net-
work is generated for the set of proteins that contains
them and their partners. In this network, a protein
that is connected to more than one “seed” is referred
as a linker-N, with N being the number of seed pro-
teins to which it is connected. Finally, proteins only
connected to one seed protein are named leafs. This
allowed us to enlarge the interaction network and de-
tect new putatively relevant proteins for the biological
pathway.
3 CONCLUSIONS
We have presented initial results on Decision Support,
Database Mining and Knowledge Discovery for In-
tracranial Aneurysms. Due to the lack of patient data,
INITIAL RESULTS ON KNOWLEDGE DISCOVERY AND DECISION SUPPORT FOR INTRACRANIAL
ANEURYSMS
269
biomedical data and compute services. IEEE Trans
Nanobioscience, 6(2):136–141.
Bonis, J., Furlong, L. I., and Sanz, F. (2006). OSIRIS: a
tool for retrieving literature about sequence variants.
Bioinformatics, pages 2567–2569.
Chatr-aryamontri, A., Ceol, A., Palazzi, L. M., Nardelli,
G., Schneider, M. V., Castagnoli, L., and Cesareni, G.
(2007). Mint: the molecular interaction database. Nu-
cleic Acids Res, 35(Database issue):D572–D574.
Clarke, M. (2007). Cochrane literature review snapshot v2.0
- aneurysm analysis rupture risk review. Deliverable
of the @neurIST project D18, University of Oxford.
Cockell, S. J., Oliva, B., and Jackson, R. M. (2007).
Structure-based evaluation of in silico predictions of
protein-protein interactions using comparative dock-
ing. Bioinformatics, 23(5):573–581.
Druzdzel, M. J. (1999). GeNIe: A development environ-
ment for graphical decision-analytic models. In Pro-
ceedings of the 1999 Annual Symposium of the Amer-
ican Medical Informatics Association (AMIA-1999),
page 1206.
Efron, B. and Tibshirani, R. J. (1993). An Introduction to
the Bootstrap. Chapman & Hall/CRC.
Espadaler, J., Romero-Isart, O., Jackson, R. M., and Oliva,
B. (2005). Prediction of protein-protein interactions
using distant conservation of sequence patterns and
structure relationships. Bioinformatics, 21(16):3360–
3368.
Fayyad, U. M., Piatetsky-Shapiro, G., and Smyth, P. (1996).
Knowledge Discovery and Data Mining: Towards a
Unifying Framework. In Knowledge Discovery and
Data Mining, pages 82–88.
Fluck, J., Mevissen, H. T., Dach, H., Oster, M., and
Hofmann-Apitius, M. (2007). ProMiner: recognition
of human gene and protein names using regularly up-
dated dictionaries. In (Hirschmann et al., 2007), pages
149–151.
Fox, J. and Das, S. (2000). Safe and sound: Artificial intel-
ligence in hazardous applications. In Proceedings of
AAAI 2000.
Fox, J., Johns, N., Rahmanzadeh, A., and Thomsen, R.
(1996). PROforma: A method and language for spec-
ifying clinical guidelines and protocols. In Brender,
J., Christensen, J., Scherrer, J.-R., and McNair, P., ed-
itors, Proceedings of the Medical Informatics Europe
1996, pages 516–520. IOS Press.
Fox, J., Patkar, V., and Thomson, R. (2006). Decision sup-
port for health care: the proforma evidence base. In-
form Prim Care, 14(1):49–54.
Gattermayer, T. (2007). SCAIView: annotation and visu-
alization system for knowledge discovery. Master’s
thesis, Life Science Informatics at Bonn-Aachen In-
ternational Center for Information Technology (B-IT);
Germany.
Ghinea, N. and van Gelder, J. M. (2004). A probabilis-
tic and interactive decision-analysis system for un-
ruptured intracranial aneurysms. Neurosurg Focus,
17(5):E9.
Glasspool, D. W., Fox, J., Oettinger, A., and Smith-Spark,
J. H. (2006). Argumentation in decision support for
medical care planning for patients and clinicians. In
Proceedings of the AAAI Spring Symposium Series
2006.
Han, B., Friedrich, C. M., Manthey, R., and Hofmann-
Apitius, M. (2006). Bayesian network modeling for
the prediction of treatment outcomes of intracranial
aneurysms. In Huson, D., Kohlbacher, O., Lupas, A.,
Nieselt, K., and Zell, A., editors, Short Papers and
Poster Abstracts of the German Conference on Bioin-
formatics (GCB2006), page 122.
Hanisch, D., Fundel, K., Mevissen, H.-T., Zimmer, R.,
and Fluck, J. (2005). ProMiner: rule-based protein
and gene entity recognition. BMC Bioinformatics, 6
(Suppl 1)(S14).
Hirschmann, L., Krallinger, M., and Valencia, A., editors
(2007). Proceedings of the Second BioCreative Chal-
lenge Evaluation Workshop. Centro Nacional de In-
vestigaciones Oncologicas, CNIO.
Hurt, C., Fox, J., Bury, J., and Saha, V. (2003). Comput-
erised advice on drug dosage decisions in childhood
leukaemia: a method and a safety strategy. In Dojat,
M., Keravnou, E., and Barahona, P., editors, Proceed-
ings of the 9th Conference on Artificial Intelligence
in Medicine in Europe (AIME’03), LNAI 2780, pages
158–163. Springer Verlag.
Jensen, L. J., Saric, J., and Bork, P. (2006). Literature min-
ing for the biologist: from information retrieval to bi-
ological discovery. Nature Reviews; Genetics, 7:119–
129.
Kerrien, S., Alam-Faruque, Y., Aranda, B., Bancarz, I.,
Bridge, A., Derow, C., Dimmer, E., Feuermann, M.,
Friedrichsen, A., Huntley, R., Kohler, C., Khadake, J.,
Leroy, C., Liban, A., Lieftink, C., Montecchi-Palazzi,
L., Orchard, S., Risse, J., Robbe, K., Roechert, B.,
Thorneycroft, D., Zhang, Y., Apweiler, R., and Her-
mjakob, H. (2007). Intact–open source resource
for molecular interaction data. Nucleic Acids Res,
35(Database issue):D561–D565.
Klinger, R., Furlong, L. I., Friedrich, C. M., Mevissen,
H. T., Fluck, J., Sanz, F., and Hofmann-Apitius, M.
(2007). Identifying gene specific variants in biomedi-
cal text. Journal of Bioinformatics and Computational
Biology, 5(6):in print.
Krischek, B. and Inoue, I. (2006). The genetics of intracra-
nial aneurysms. J Hum Genet, 51(7):587–594.
Kullback, S. and Leibler, R. (1951). On information and
sufficiency. Ann. Math. Stat., 22:79–86.
Lafferty, J., McCallum, A., and Pereira, F. C. N. (2001).
Conditional Random Fields: Probabilistic Models for
Segmenting and Labeling Sequence Data. In Pro-
ceedings of the Eighteenth International Conference
on Machine Learning (ICML 2001), pages 282–289.
Morgan Kaufmann Publishers.
Maglott, D., Ostell, J., Pruitt, K. D., and Tatusova, T.
(2005). Entrez Gene: gene-centered information at
NCBI. Nucleic Acids Research, 0:D1–D6.
INITIAL RESULTS ON KNOWLEDGE DISCOVERY AND DECISION SUPPORT FOR INTRACRANIAL
ANEURYSMS
271
Molyneux, A. J., Kerr, R. S. C., Yu, L., Clarke, M., Sneade,
M., Yarnold, J. A., and Sandercock, P. (2005). Inter-
national subarachnoid aneurysm trial (ISAT) of neuro-
surgical clipping versus endovascular coiling in 2143
patients with ruptured intracranial aneurysms: a ran-
domised comparison of effects on survival, depen-
dency, seizures, rebleeding, subgroups, and aneurysm
occlusion. Lancet, 366:809–817.
Morgan, A. A. and Hirschmann, L. (2007). Overview of
BioCreative II gene normalization. In (Hirschmann
et al., 2007), pages 17–27.
Nahed, B. V., Bydon, M., Ozturk, A. K., Bilguvar, K.,
Bayrakli, F., and Gunel, M. (2007). Genetics of in-
tracranial aneurysms. Neurosurgery, 60(2):213–25;
discussion 225–6.
Pagel, P., Kovac, S., Oesterheld, M., Brauner, B., Dunger-
Kaltenbach, I., Frishman, G., Montrone, C., Mark, P.,
St
¨
umpflen, V., Mewes, H.-W., Ruepp, A., and Frish-
man, D. (2005). The mips mammalian protein-protein
interaction database. Bioinformatics, 21(6):832–834.
Patkar, V., Hurt, C., Steele, R., Love, S., Purushotham,
A., Williams, M., Thomson, R., and Fox, J. (2006).
Evidence-based guidelines and decision support ser-
vices: A discussion and evaluation in triple assess-
ment of suspected breast cancer. Br J Cancer,
95(11):1490–1496.
Peri, S., Navarro, J. D., Amanchy, R., Kristiansen, T. Z.,
Jonnalagadda, C. K., Surendranath, V., Niranjan, V.,
Muthusamy, B., Gandhi, T. K. B., Gronborg, M., Ibar-
rola, N., Deshpande, N., Shanker, K., Shivashankar,
H. N., Rashmi, B. P., Ramya, M. A., Zhao, Z., Chan-
drika, K. N., Padma, N., Harsha, H. C., Yatish, A. J.,
Kavitha, M. P., Menezes, M., Choudhury, D. R.,
Suresh, S., Ghosh, N., Saravana, R., Chandran, S., Kr-
ishna, S., Joy, M., Anand, S. K., Madavan, V., Joseph,
A., Wong, G. W., Schiemann, W. P., Constantinescu,
S. N., Huang, L., Khosravi-Far, R., Steen, H., Tewari,
M., Ghaffari, S., Blobe, G. C., Dang, C. V., Garcia,
J. G. N., Pevsner, J., Jensen, O. N., Roepstorff, P.,
Deshpande, K. S., Chinnaiyan, A. M., Hamosh, A.,
Chakravarti, A., and Pandey, A. (2003). Develop-
ment of human protein reference database as an initial
platform for approaching systems biology in humans.
Genome Res, 13(10):2363–2371.
Rinkel, G. J., Djibuti, M., Algra, A., and van Gijn, J.
(1998). Prevalence and risk of rupture of intracranial
aneurysms. Stroke, 29:251–256.
Risselada, R., van der Lugt, A., Niessen, W. J., and Sturken-
boom, M. C. J. M. (2007). Hormonal contraceptives
and risk of aneurysmatic subarachnoid haemorrhage.
Basic & Clinical Pharmacology & Toxicology, Spe-
cial Issue on 8th Congress of the European Associa-
tion for Clinical Pharmacology and Therapeutics, 101
(Suppl.):40.
Rual, J.-F., Venkatesan, K., Hao, T., Hirozane-Kishikawa,
T., Dricot, A., Li, N., Berriz, G. F., Gibbons, F. D.,
Dreze, M., Ayivi-Guedehoussou, N., Klitgord, N., Si-
mon, C., Boxem, M., Milstein, S., Rosenberg, J.,
Goldberg, D. S., Zhang, L. V., Wong, S. L., Franklin,
G., Li, S., Albala, J. S., Lim, J., Fraughton, C., Llam-
osas, E., Cevik, S., Bex, C., Lamesch, P., Sikorski,
R. S., Vandenhaute, J., Zoghbi, H. Y., Smolyar, A.,
Bosak, S., Sequerra, R., Doucette-Stamm, L., Cusick,
M. E., Hill, D. E., Roth, F. P., and Vidal, M. (2005).
Towards a proteome-scale map of the human protein-
protein interaction network. Nature, 437(7062):1173–
1178.
Russel, S. and Norvig, P. (2003). Artificial Intelligence a
modern Approach. Prentince Hall, 2nd edition.
Salwinski, L., Miller, C. S., Smith, A. J., Pettit, F. K.,
Bowie, J. U., and Eisenberg, D. (2004). The database
of interacting proteins: 2004 update. Nucleic Acids
Res, 32(Database issue):D449–D451.
Smigielski, E. M., Sirotkin, K., Ward, M., and Sherry, S. T.
(2000). dbSNP: a database of single nucleotide poly-
morphisms. Nucleic Acids Research, 28(1):352–355.
Stark, C., Breitkreutz, B.-J., Reguly, T., Boucher, L., Bre-
itkreutz, A., and Tyers, M. (2006). Biogrid: a general
repository for interaction datasets. Nucleic Acids Res,
34(Database issue):D535–D539.
Stelzl, U., Worm, U., Lalowski, M., Haenig, C., Brem-
beck, F. H., Goehler, H., Stroedicke, M., Zenkner, M.,
Schoenherr, A., Koeppen, S., Timm, J., Mintzlaff, S.,
Abraham, C., Bock, N., Kietzmann, S., Goedde, A.,
Toks
¨
oz, E., Droege, A., Krobitsch, S., Korn, B., Birch-
meier, W., Lehrach, H., and Wanker, E. E. (2005). A
human protein-protein interaction network: a resource
for annotating the proteome. Cell, 122(6):957–968.
Sutton, D. R. and Fox, J. (2003). The syntax and semantics
of the proforma guideline modeling language. J Am
Med Inform Assoc, 10(5):433–443.
van der Lei, J., Duisterhout, J. S., Westerhof, H. P., van der
Does, E., Cromme, P. V., Boon, W. M., and van Bem-
mel, J. H. (1993). The introduction of computer-based
patient records in the netherlands. Ann Intern Med,
119(10):1036–1041.
von Mering, C., Jensen, L. J., Snel, B., Hooper, S. D.,
Krupp, M., Foglierini, M., Jouffre, N., Huynen,
M. A., and Bork, P. (2005). String: known and
predicted protein-protein associations, integrated and
transferred across organisms. Nucleic Acids Res,
33(Database issue):D433–D437.
Wiebers, D., Whisnant, J., Huston, J., Meissner, I.,
Brown RD, J., Piepgras, G., Thielen, K., Nichols,
D., O’Fallon, W., Peacock, J., Jaeger, L., Kassell, N.,
Kongable-Beckman, G., and Torner, J. (2003). Unrup-
tured intracranial aneurysms: natural history, clinical
outcome, and risks of surgical and endovascular treat-
ment. Lancet, 362:103–110.
HEALTHINF 2008 - International Conference on Health Informatics
272