Lowering the barriers to cancer imaging

Lowering the barriers to cancer imaging

Lowering the barriers to cancer imaging is part of a Cancer Imaging project funded by the Technical Initiative of Microsoft Corporation.

  • Provide a framework for medical image analysis (MIA) researchers to share algorithms and image data with other MIA researchers and clinicians.
  • Investigate the use of  mutli-touch and mutli-user technology for  a more collaborative environment for the analysis of medical images.

Approximately 36,000 people are diagnosed with colorectal cancer every year in UK which makes it the third most common cancer in UK. Furthermore, colorectal cancer often metastasizes to the liver with poor prognosis, and liver cancer itself causes around 3,000 deaths each year in the UK. 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 with colorectal and liver cancer. Radiologists analyze medical images to detect abnormalities, and when one of these is categorized as a tumor it has to be characterized; its size, location and configuration provide information to support the prognosis.

Image segmentation and registration are keys to such a categorization and to support the decision-making for treatment delivery and the response analysis by comparing tumor shape, location and volume at different time periods. However, colorectal cancer images are often noisy, complex and highly textured and the segmentation of tumors is challenging due to the poor contrast relative to their surroundings.

The types of segmentation and registration algorithms that are suitable for any given image will depend on several parameters, and researchers are continually working to improve on existing algorithms. When a new project starts a researcher must consider all existing solutions to determine if they are suitable or not. This may be very time consuming and may not lead to relevant algorithms that are implemented in a way that the researcher can adopt the solution. On the other hand, clinicians need to become familiar with the software solutions developed by MIA researchers in order to provide feedback about their performance as well as to generate data for validating the results obtained.

This project will explore some of the difficulties that medical image analysis researchers and clinicians face and is developing an approach to improve their efficiency capability using innovative computational tools.


Avila-Garcia, M.S.   Trefethen, A.E.   Brady, M.   Gleeson, F.   Goodman, D.
Lowering the Barriers to Cancer Imaging. IEEE Fourth International Conference on eScience, pp 6370, 2008. ISBN: 978-1-4244-3380-3.

Case Studies:

Microsoft Case Study


Oxford e-Research Centre

Medical Vision Laboratory

Churchill Hospital

  • Dr. Fergus Gleeson.


  • Dr Maria Susana Avila Garcia: susana.garcia {at} oerc.ox.ac.uk