Medical Imaging and Image Interpretation Group

Note: this page is under construction. An updated version of the group webpage will be available soon.

The Medical Imaging and Image Interpretation Group is focused on finding novel, non-invasive methods of acquiring image data and extracting information from these data. Some of the recent work has been focused on developing novel techniques that allow us to use the features and properties of an optical image to infer quantitative information about the structure and composition of the tissue being imaged. Specific areas of work include:

  • Optical imaging for the detection of skin cancer: This work relies on a detailed understanding of the physics of image formation, and has led to the development of the successful skin imaging system, the SIAScope.
  • Detection of early signs of retinopathies: A recently completed project has successfully demonstrated the ability of these methods to detect and quantify retinopathies from images of the human eye.
  • Optical diagnosis of colon cancer.
  • Detection and characterisation of breast cancer using optical imaging.
  • Optical tomographic imaging of the human brain.
  • Small animal imaging using molecular markers.
  • Multi-modality imaging and image coregistration.

We are also developing new techniques for modelling light transport in tissues. A finite element model based image reconstruction package (NIRFAST), freely available and developed for MATLAB, has been developed and is being used for applications in optical imaging using intrinsic and extrinsic molecular markers. We are also especially interested in the application of cutting-edge techniques from natural computation to help us to explore and optimise our models and imaging methods. The group is also interested in statistical methods of image analysis, especially the identification of the boundaries of lesions, and the extraction of textural information from an image. This information can be of great benefit to a clinician when making a diagnosis.

The group hosts and maintains a fully-equipped multispectral imaging laboratory, part of the university's Collaborative Research Network in Imaging and Visualisation. Further information about the group can be found here.