Transcript D2_4_1515_Thakar_en

DICOM – A Preclinical Perspective
AK Narayan, Kishan Harwalkar, Kshitija Thakar
Philips Healthcare,
April 09, 2008
Agenda
 Introduction to Preclinical
 IMALYTICS Workspace
 Information Model Requirements
 Mapping to DICOM
 DICOM Constraints on Preclinical
 Challenges and Future Work
 Conclusions
CONFIDENTIAL
Philips Healthcare, Apr 09, 2008
2
Research Workflow
 Research is characterized by exploratory and/or hypothesis-driven
programs often supported by grants to either discover or explore new
insights into biological processes.
 The systematic discovery and development of biomarkers, drugs, and
therapies that will ultimately be translated from animal models to human
should they prove promising during preclinical studies.
Exploratory
or Hypothesis
driven
CONFIDENTIAL
Transition to
...
Statistically
Significant
Results
Enabled by
Pre-Clinical
Workspace
Philips Healthcare, Apr 09, 2008
3
Preclinical Imaging
“Researchers are not mouse doctors”
Fundamental
Understanding
of Biology/
Biochemistry
• transfer in-vitro
to in-vivo
• verify models
Massoud T.F., Gambhir S.S.;
Molecular Imaging in living
subjects: seeing fundamental
biological processes in a new
light; Genes Dev., 17, 545-580,
2003
CONFIDENTIAL
Design and
Evaluation of new
Biomarkers (drugs)
(diagnosis/therapy)
• dynamics and kinetics
• efficacy (candidate
selection)
• dosing
Rudin M., Weissleder R.;
Molecular Imaging in drug
discovery and development
Nat. Rev. Drug Discov., 2, 123131, 2003
Philips Healthcare, Apr 09, 2008
Test Bed for new
Imaging
Technologies
• small size prototypes
• low capital investment
• POC (proof of concept)
Gleich B., Weizenecker J.;
Tomographic Imaging using
the nonlinear response of
magnetic particles
Nat., 435(30), 1214-1217, 2005
4
Imaging applications in drug discovery
and development
Rudin M., Weissleder R.;
Molecular Imaging in drug discovery and development
Nat. Rev. Drug Discov., 2, 123-131, 2003
CONFIDENTIAL
Philips Healthcare, Apr 09, 2008
5
Multi-modality in Preclinical
Massoud & Gambhir, Genes & Development, 2003
CONFIDENTIAL
Philips Healthcare, Apr 09, 2008
6
Preclinical application needs
Increasing the productivity, reproducibility, and standardization of a
variety of experimental approaches such as:
– Snapshot measurement on a single subject
– Longitudinal studies on the single subject across multiple
sessions
– Group studies on multiple subjects in the same laboratory
– Studies on distributed population groups that are done to
substantiate the hypothesis.
CONFIDENTIAL
Philips Healthcare, Apr 09, 2008
7
IMALYTICS Workspace
 Multi-modality Preclinical Workstation
 Provides a combined view of the different facets of the drug discovery
process.
 Provides advanced image analysis, quantification, and visualization
tools dedicated to research and discovery
CONFIDENTIAL
Philips Healthcare, Apr 09, 2008
8
IMALYTICS Modeling requirements
 Data Mining
– Project Oriented View
– Each Preclinical Project will involve multiple Subjects with Series of
images under each
 Interoperability
– High Interoperability with existing Standards
– High Interoperability with existing Preclinical data
 Compatibility
– Extendable for Clinical Trials
– Compatible with existing Clinical Apps
CONFIDENTIAL
Philips Healthcare, Apr 09, 2008
9
Preclinical Real-World Model
Project
Project ID
Description
Principal Investigator
1..n
1..n
Subject
Subject ID
Strain Name
Sex
Series
Series Number
Modality
Series Description
1..n
1..n
Study
Study ID
Study Date
CONFIDENTIAL
Image
SOP Common Module
Image Pixel Module
Philips Healthcare, Apr 09, 2008
1..n
Non Image
SOP Common Module
10
DICOM Clinical Trial Model
Patient
Clinical Trial Subject
1..n
Patient ID
Patient Name
Patient Sex
Clinical Trial Sponsor Name
Clinical Trial Protocol Name
Clinical Trial Protocol ID
1..n
Study
Study ID
Study Date
1..n
Series
1..n
Series Number
Modality
Series Description
Image
SOP Common Module
Image Pixel Module
CONFIDENTIAL
Philips Healthcare, Apr 09, 2008
11
Mapping Preclinical Model to DICOM
Patient

Depicts the Conceptual
Model of the Preclinical
domain

Data mining at Project Level
would be easy
Study

Less Compatibility with
existing clinical applications
Series

Low Interoperability with
clinical DICOM data
Project
Clinical Trial Subject
Subject
Study
Series
Image
CONFIDENTIAL
Image
Philips Healthcare, Apr 09, 2008
12
Mapping Preclinical Model to DICOM
Project
Patient
Clinical Trial Subject
Subject
Study
Study
Series
Image
CONFIDENTIAL
Series
Image
Philips Healthcare, Apr 09, 2008

High co-relation with DICOM
Model

High Compatibility with
existing Clinical Applications

High Interoperability

Differs from the Conceptual
Preclinical model

Data mining at Project level
is not easy
13
IMALYTICS Model
Patient
Project
Interoperability :
 Easily achievable
Subject
Compatibility :
 Can be used for Clinical Trial
 Clinical Apps can be easily
integrated
Clinical Trial Subject
Study
Series
Image
CONFIDENTIAL
Study
Data Mining :
 Not possible via the Model
 Can be achieved via Software
Series
Semantic Correlation :
 High Correlation with DICOM
Image
Philips Healthcare, Apr 09, 2008
14
Project-oriented Workflow
CONFIDENTIAL
Philips Healthcare, Apr 09, 2008

Enables Project Oriented
View with local Database

The Project Oriented view
can be seamlessly used for
Clinical Trials

Project Oriented View may
not be possible with DICOM
Network and Media

Importing a Complete
Project in one shot is not
possible
15
DICOM Constraints on Preclinical
Group Studies
Multiple Subjects are a part of the
same Scan
– Sharing the same Study after
splitting into multiple
hierarchy is not possible
– Orientation of individual
subject can not be
represented
CONFIDENTIAL
Philips Healthcare, Apr 09, 2008
16
DICOM Constraints on Preclinical
 DICOM Type 2 attributes may not always be applicable in the Preclinical
domain
– Patient Birth Date
– Patient Sex
– Referring Physician
CONFIDENTIAL
Philips Healthcare, Apr 09, 2008
17
Challenges and Future Work
• Challenges
– IMALYTICS Model vs. Preclinical model by other vendors
– Non availability of Data from different vendors
– Non availability of DICOM Conformance Statements for preclinical
products
• Future Work
– Extending the IMALYTICS model for Group Studies
– Applicability of the current model for
• Distributed population Study
• Clinical Trials
– Dealing with non-image data like Histology (in-silico, in-vitro, exvivo)
CONFIDENTIAL
Philips Healthcare, Apr 09, 2008
18
Conclusions
 Preclinical Imaging has emerged recently as a powerful tool that
enables Clinical Research
 Going forward Interoperability would be the key in Preclinical domain
(especially for translational research)
 Specific platforms to address Interoperability in Preclinical
(IHE/Connectathon) are required
CONFIDENTIAL
Philips Healthcare, Apr 09, 2008
19