University College London

UCL has a global reputation for excellence in research and is committed to delivering impact and innovations that enhance the lives of people in the UK, across Europe and around the world. UCL is consistently placed in the global top 20 across a wide range of university rankings (currently 4th in the QS World University Rankings with a score of 98.9). Furthermore, the Thomson Scientific Citation Index shows that UCL is the 2nd most highly cited European university and 14th in the world.

UCL’s total competitively awarded research income annually stands at an impressive € 530 million, of which 10% is European funded research & innovation. UCL is one of the leading recipients of European Framework Programme grants, with over 600 projects funded under the Seventh Framework Programme (FP7) including more than 100 prestigious European Research Council awards. In line with UCL’s position as Europe’s 6th most active Higher Education organisation in FP7, UCL has expanded its European Research and Innovation Office which is currently managing a total of 24 concurrent FP7 projects with a total combined budget of over €148 million.

The Centre for Medical Image Computing (CMIC) (Department of Computer Science, UCL) was established in 2005 to bring together imaging researchers in computer science and medical physics with users in medicine and life sciences. CMIC plays a key role in translating to biomedical sciences the new developments in imaging methodology. It includes seven academic staff, over 60 research staff and students, and a diverse group of collaborators. CMIC’s income since 2005 is in excess of £20M.

The UCL group is part of a large centre (CMIC) specialising in the modelling, reconstruction and analysis of medical images from all modalities, including MRI, CT, PET, SPECT, Ultrasound and Optics. In addition the Centre for Inverse Problems has extensive experience of combining mathematical, statistical, and computational techniques for forward and inverse problems arising in imaging.

UCL’s involvement will be in Work Package 2 Modelling, computation and data processing, which it will lead. UCL will develop efficient models of light propagation in diffusive media, tailored for the novel photonic module design in the SOLUS project. UCL will provide tomographic reconstruction algorithms suitable for the developed instrumentation and their implementation on state-of-the-art GPU and distributed architectures.  UCL will analyse the modelling and reconstruction problems to propose an optimal design for the experimental system in terms of information content of the reconstructed images. UCL will use methods for joint reconstruction to optimize the simultaneous recovery of optical and US/elastography quantitative images.

A key novel aspect of SOLUS is the classification of lesions using multimodality information. UCL will use techniques from Machine Learning to iteratively optimize reconstruction and classification. This will build on prior work applying these methods to multispectral optical, photoacoustic and electrical impendance tomographic images. In this approach the posterior classification confidence estimates are optimized with the constraint that the reconstructed images are consistent with the measured data with noise limits of the experimental system.

For more information go to ucl.ac.uk