Science Applications Supporting the UK tomography community with a toolbox of algorithms to increase the quality and level of information that can be extracted by computed tomography. Enhancing the UK's global leadership in the field of medical imaging by positron emission tomography (PET) and magnetic resonance (MR) imaging by bringing together the best of the UK's imaging expertise to capitalise on the investment in this area. For example, community-run hackathon training sessions demonstrated that their tools allow biomedical researchers to implement a new reconstruction algorithm and test it on real scanner data within days, rather than over months even with advanced programming skills. The acceleration of the development of reconstruction algorithms will translate into faster progress in accuracy of scanners' images, which will lead to better quality and longevity of life for people affected by cancer, dementia and other serious illnesses.
Expertise Collaborative research is achieved via software and methodology development, training and outreach activities, dissemination, community building, networking and providing access to sustained networks of communities.
More specifically, core computational support includes: developing, maintaining, sustaining, software engineering, testing, deploying and documenting the core imaging toolbox and the synergistic imaging reconstruction framework. Reducing the barrier to access in multi-modality image analysis algorithms; improving the accessibility and distribution of the codes; establishing a national multidisciplinary image analysis focal point for the multidisciplinary community comprising algorithm developers, material scientists, instrument manufacturers, and instrument scientists.
Software Knowledgeable and experienced CoSeC staff within the Tomographic and Biological Imaging communities support and/or develop the Core Imaging Library (CIL), and the Synergistic Image Reconstruction Framework (SIRF) code base as tabulated below. Additionally, the CCPi software suite includes algorithms for pre-processing, quantification, segmentation, and visualisation, and methods that are resilient to low signal-to-noise-ratios; the Stochastic Primal Dual Hybrid Gradient (SPDHG) algorithm for minimisation, readers for data acquired by NIKON and ZEISS machines, tools for pre-processing data, e.g. normalisation, transmission to absorption, and a plugin for the TIGRE engine.
CCPs and Project Leads
| |
