Dr. Maria Susana Avila Garcia1, Prof Anne E. Trefethen1, Prof Sir Michael Brady2, Dr Fergus Gleeson3 and Dr. Daniel Goodman1

1
Cloud Computing Framework Design for Cancer
Imaging Research

• Dr. Maria Susana Avila Garcia1, Prof Anne E.
  Trefethen1, Prof Sir Michael Brady2, Dr Fergus
  Gleeson3 and Dr. Daniel Goodman1

1. Oxford e-Research Centre, University of
Oxford, UK 2. Dept. of Eng. Science, University
of Oxford, UK 3. Radiology, Nuffield Dept. of
Medicine, Churchill Hospital, University of
Oxford, UK
2
Outline

• Colorectal Cancer
• Oxford approach
• Cancer and Cardiac Imaging Project
• Lowering the Barrier to Cancer Imaging
• Cloud Computing Framework
• Microsoft Tools
• Challenges
• Future Work
• Conclusions

3
Colorectal and liver cancer in UK

• According with Cancer Research UK (cited August
  2008)
• Approximately 36,000 people are diagnosed with
  colorectal cancer every year in UK
• The third most common cancer
• Colorectal cancer often metastasizes to the liver
  with poor prognosis,
• liver cancer causes around 3,000 deaths each
  year.
• Medical imaging techniques such as magnetic
  resonance imaging (MRI), ultrasound (US),
  computerized tomography (CT) and a combination of
  positron emission tomography (PET) with CT
  (PET/CT), have been used for detecting, staging,
  and monitoring the evolution of patients

4
At Oxford

• Researchers working in image analysis of
  colorectal and liver cancer images
• Segmentation
• Registration
• Image quality improvement.

• Analysis of medical images is difficult since
  they are
• Noisy,
• Highly textured,
• Poor contrast relative to their surroundings.

Coronal MR image of the colorectum
5
Cancer and Cardiac Imaging

• Technical Computing Initiative project funded by
  Microsoft Corporation
• Investigating the development of new segmentation
  algorithms for colorectal cancer imaging. Dr.
  Niranjan Joshi and Prof Sir Mike Brady (OERC)
  (Engineering Department Oxford University) and
  Dr. Fergus Gleeson (Churchill Hospital and Oxford
  University) , and Prof. Andrew Blake (Microsoft
  Research Cambridge)
• Lowering the Barriers to Cancer Imaging project
  is aimed to maximise the efficiency of a Medical
  Image Analysis (MIA) researcher and to alleviate
  the frustration of clinicians for not being able
  to analyse and process images using the
  algorithms developed by MIA researchers. PIs
  Prof Anne E. Trefethen and Prof Sir Mike Brady
  (OERC)

6
Lowering the Barriers to Cancer Imaging

• SHARING RESOURCES
• A platform independent framework.
• Federated storage (data, algorithms, related
  info).
• A repository of algorithms with no bounds to
  specific programming languages.
• Access to already existing imaging and
  visualization toolkits with no bounds to specific
  programming languages.
• Access to the most up-to-date authoritative
  knowledge.
• A framework for rapid development and deployment
  of applications for use by researchers and
  clinicians.
• Improve mechanisms for manual segmentation

7
Lowering the Barriers to Cancer Imaging

• APPLICATION DESIGN
• Use of Collaborative visual tools (including
  multi-touch and interactive surfaces) to improve
  visual data input and enhance user interaction.

8
Cloud Computing Framework
Provenance contributions of each researcher are
registered and the use of their methods and
experimental data is acknowledged
9
Cancer Imaging Cloud Computing Framework
Image processing Visualization toolkits
User interface tools
Workflows
Metadata
Web Services
My Experiment Carmen Research Information Centre,
RIC
SciRun IRIS Explorer Matlab
Taverna Microsoft Workflow Foundation
10
Microsoft Tools

• Visual Studio is being already used by MIA
  researchers and makes it easy to add Web Service
  calls.
• Use .NET platform to develop application to
  enable the use of a unique platform
• Including Microsoft Workflow Foundation.
• Collaborate with existing Virtual Research
  Environments
• Research Information Centre (RIC)

11
Challenges

• The adaptation of existing software
• Virtual research environments.
• Imaging and Visualisation toolkits.
• Algorithms developed by researchers.
• Link to permanent and secure online archives,
• Repository for research materials produced by
  scholars at Oxford University, to ensure access
  to a permanent and secure online archive,
  http//ora.ouls.ox.ac.uk/
• Repositories with Cancer Images, i.e. National
  Cancer Imaging Archive (NCIA). https//imaging.nci
  .nih.gov/ncia/

12
Challenges

• Engage potential users
• Medical image analysis (MIA) researchers
• define the way contribution will be made.
• Engineering and computer science academics, and
  to undergraduate students,
• to raise interest in challenges to solve
  computational and software engineering problems.
• Engage medical and biomedical science academics
  and students with the use of image processing
  techniques

13
Future work
14
Conclusions

• We have presented a Cloud computing framework
  design to provide
• Rapid application testing and development
  environment for Medical Image Analysis (MIA)
  researchers.
• Easy access to federated resources (algorithms
  and data) for both MIA researchers and
  Clinicians.
• Support to imaging and visualization toolkits
  using Visual Studio.
• We have outlined our plan for future work which
  includes collaboration with other projects.

15
Acknowledgements

• This research is funded by the Technical
  Computing Initiative of Microsoft Corporation.
• We thank MIA researchers at Oxford for their
  valuable comments during the analysis of
  requirements for this project, especially Vicente
  Grau, Niranjan Joshi and Olivier Noterdaeme as
  well as radiologists working at Churchill and
  John Radcliffe Hospitals especially Dr. Rachel R
  Phillips and Dr Mark Anderson.