HealtheChild

1
Health-e-Child
Peter Bloodsworth CCCS Research Centre UWE,
Bristol, UK peter.bloodsworth_at_cern.ch
HP Labs Presentation 23th November 2007
2
Talk Structure

• The Health-e-Child Project.
• The Semantic Web in Health-e-Child.
• Past Research and Future Projects.
• NeuGrid.
• Conclusion.

HP Labs Presentation 23th November 2007
3
Facts and Figures

• Coordinator Siemens AG
• Partner 14 European companies, hospitals,
  institutions
• Timetable 01-Jan-06 to 31-Dec-09 (4 years)?
• Web page http//www.Health-e-Child.org
• Instrument Integrated Project (IP) of the
• Framework Program FP6
• Project Identifier IST-2004-027749

HP Labs Presentation 23th November 2007
4
Clinical Aspects Focus on Paediatrics

• Diseases
• Heart diseases (Right Ventricle Overload,
  Cardiomyopathy.)?
• Inflammatory diseases (Juvenile Idiopathic
  Arthritis.)?
• Brain tumours (Gliomas.)?
• Clinical Institutions
• I.R.C.C.S. Giannina Gaslini, Genoa, Italy.
• University College London, Great Ormond Street
  ChildrensHospital, London, UK.
• Assistance Publique Hopitaux de Paris Necker,
  Paris, France.

HP Labs Presentation 23th November 2007
5
The Health-e-Child Consortium

• Siemens AG, Germany
• Lynkeus SRL, Rome, Italy
• Giannina Gaslini Hospital, Genoa, Italy (IGG)?
• Great Ormond Street Childrens Hospital, UK
• Assistance Publique Hopitaux de Paris Necker,
  Paris, France (APHP)?
• European Organisation for Nuclear Research
  (CERN), Geneva, Switzerland
• Maat G Knowledge, Toledo, Spain
• University of the West of England, Bristol, UK
• University of Athens, Athens, Greece (UoA)?
• Universita' degli Studi di Genova (DISI),
  Genoa, Italy
• National Institute for Information and
  Automation Research (INRIA), Sophia
• Antipolis, France
• European Genetics Foundation (EGF), Bologna,
  Italy
• Aktsiaselts ASPER BIOTECH, Tartu, Estonia
• Gerolamo Gaslini Foundation, Genoa, Italy (FGG)?

HP Labs Presentation 23th November 2007
6
Juvenile Idiopathic Arthritis

• Juvenile idiopathic arthritis is a subset of
  arthritis seen in childhood.
• May be transient and self-limited or chronic.
• It differs significantly from arthritis commonly
  seen in adults (osteoarthritis, rheumatoid
  arthritis.)?
• JIA is a fairly rare disease lt 1 in 1000 children
  suffer from it.
• It is difficult to diagnose because of the range
  of symptoms.
• Swelling may be difficult to detect clinically.
• Young children may have difficulty in
  communicating pain.

Source Wikipedia
HP Labs Presentation 23th November 2007
7
Project Aims

• To gain a holistic view of a childs health by
  integrating traditional and emerging sources of
  biomedical information from genetic to clinical
  to epidemiological across institutions
  (horizontal, vertical, and longitudinal
  integration.)?
• Develop a biomedical information platform,
  supported by sophisticated and robust search,
  optimisation, and matching techniques for
  heterogeneous information, empowered by the Grid.
• Build enabling tools and services that will
  improve the quality of care and reduce its cost
  by increasing efficiency.

HP Labs Presentation 23th November 2007
8
Health-e-Child
HP Labs Presentation 23th November 2007
9
Project Outcomes

• Outcomes of Health-e-Child include
• The construction of Integrated disease models
  exploiting all available information levels.
• Database-guided biomedical decision support
  systems.
• ? Large-scale, cross modality information fusion
  and data mining for knowledge discovery.
• The use of Semantic Web technologies such as
  ontologies to capture domain knowledge and assist
  in providing services to clinicians.

HP Labs Presentation 23th November 2007
10
The Role of the Semantic Web in Health-e-Child

• Health-e-Child is a data integration project.
• The Semantic Web is used in many parts of the
  platform
• To power vertical data integration.
• Providing domain knowledge to decision support
  systems.
• Assisting users during querying and interaction
  with the platform.
• Re-use of existing medical domain knowledge by
  the extraction and linkage of ontological
  fragments.
• To help us manage the heterogeneity and general
  complexity of medical records.
• Standardisation of medical terminology.

HP Labs Presentation 23th November 2007
11
Case Study Assisting Users During Query
Formulation

• Example Dr Jones wants to study patients who
  are suspected of having a certain Brain Disease,
  show all patients who are likely to have Brain
  Tumour Disease-X.
• SQL for this query might be SELECT FROM
  PATIENT_INFORMATION WHERE PATIENT_ID IN (SELECT
  DISTINCT PATIENT_ID FROM PATIENT_INFORMATION X
  WHERE NOT EXISTS (SELECT FROM (SELECT CT.NAME
  CLINICAL_TEST_NAME,CTV.CT_VALUE
  CLINICAL_TEST_VALUE FROM CLINICAL_TEST CT JOIN
  CLINICAL_TEST_VALUES CTV ON CT.IDCTV.CT_ID WHERE
  (CT.NAME 'DOUBLE_VISION' AND CTV.CT_VALUE
  'TRUE') OR (CT.NAME 'HEADACHES' AND
  CTV.CT_VALUE 'TRUE') OR (CT.NAME
  'ORTHOPEDIC_SEQUELEA' AND CTV.CT_VALUE
  'SEVERE_SYMPTOMATIC')) S WHERE NOT EXISTS (SELECT
  FROM PATIENT_INFORMATION Z WHERE
  (Z.PATIENT_IDX.PATIENT_ID) AND
  (Z.CLINICAL_TEST_NAMES.CLINICAL_TEST_NAME)
  AND(Z.CLINICAL_TEST_VALUE S.CLINICAL_TEST_VALUE
  ))))
• This is a basic example things can get far more
  complicated!!

HP Labs Presentation 23th November 2007
12
The Problem of Querying

• Clinicians role is to save lives and not spend
  their time learning SQL!
• There is a range of types of queries that users
  want to run generally these are fairly complex.
• Complexity is increased by security concerns,
  data is held at individual hospitals and
  therefore we need to overlay semantic description
  on the existing data.
• Domain information is often required to enrich
  queries for example If a clinician requests a
  particular organ it might be useful to search for
  both the organ and related tissues.
• But domain information is complex itself so how
  do we use it?

HP Labs Presentation 23th November 2007
13
Can Ontological Description Help Clinicians Out?

• We can describe user queries in our ontology
  showing what data is required to produce an
  answer.
• The underlying data structure and data model and
  mapping with query groups can also be expressed
  in the ontology.
• Fragments of medical ontologies can be added to
  provide domain specific knowledge.
• A query formulation engine could use all this
  information to generate SQL that can be run
  against distributed databases and the results
  integrated to build an answer set.
• Another benefit of this approach is that the
  knowledge can also be used to help clinicians
  interact with the system.

HP Labs Presentation 23th November 2007
14
An Ontology-based Query Formulation System
HP Labs Presentation 23th November 2007
15
An Ontology Fragment
HP Labs Presentation 23th November 2007
16
Capturing DB metadata in an Ontology
HP Labs Presentation 23th November 2007
17
Future Work

• To fully populate the query formulation
  algorithms.
• Exploring further uses of ontologies to assist
  users.
• The integration of reasoners within the query
  formulation process.
• Finding ways for clinicians to extend the system
  by adding new concepts into our ontology.

HP Labs Presentation 23th November 2007
18
Past Projects and the Future

HP Labs Presentation 23th November 2007
19
The neuGrid Consortium

HP Labs Presentation 23th November 2007
20
Project Objectives

• To build a new user-friendly Grid-based research
  e-Infrastructure.
• Collection/archiving of large amounts of imaging
  data.
• Paired with computationally intensive data
  analyses.
• To enable EU neuroscientists to carry out
  cutting-edge research.
• Imaging of degenerative brain diseases.
• Google for Brain Imaging.

HP Labs Presentation 23th November 2007
21
Question Time !!

None like this please!!
HP Labs Presentation 23th November 2007