Jonathan Mason
Tue 29 Aug 2017, 14:45 - 15:00
AGB Seminar Room AGB Building, King’s Buildings, EH9 3JL

If you have a question about this talk, please contact: Ardimas Purwita (s1600157)

Image for A Monte Carlo Framework for Low Dose CT Reconstruction Testing

Jonathan Mason is currently in his third year of PhD studies at the University of Edinburgh

Abstract:
We present a framework using freely available tools for the  synthesis of physically realistic computed tomography (CT) measurements  for low dose reconstruction development and validation, using a fully  sampled Monte Carlo method. This allows the generation of test data that  has artefacts such as photon starvation, beam-hardening and scatter,  that are both physically realistic and not unfairly biased towards  model-based iterative reconstruction (MBIR) algorithms. Using the open  source Monte Carlo tool GATE and spectrum simulator SpekCalc, we  describe how physical elements such as source, specimen and detector may  be modelled, and demonstrate the construction of fan-beam and cone-beam  CT systems. We then show how this data may be consolidated and used with  image reconstruction tools. We give examples with a low dose  polyenergetic source, and quantitatively analyse reconstructions against  the numerical ground-truth for MBIR with simulated and `inverse crime'  data. The proposed framework offers a flexible and easily reproducible  tool to aid MBIR development, and may reduce the gap between synthetic  and clinical results.

Biography:
Jonathan Mason received a Master's degree (MEng.) in Electrical and  Electronic Engineering from Imperial College London in 2014, and spent  his final year at ETH Zurich with a thesis on sparsity in computed  tomography. He is currently in his third year of PhD studies at the  University of Edinburgh investigating quantitative X-ray imaging for  radiotherapy, under the supervision of Prof. Mike Davies and Dr. Bill  Nailon.