Updated: November 2014
Places in Computer Science
- Physics simulation for real-time games. Sponsor EA Ltd (Jan 2015)
- Computational Models of Argument for Supporting Medical Decision Making. Sponsored by Royal Free Charity (Jan 2015)
|< Prev||Next >|
These adverts listed in this section are all now closed and are listed to reference the scope of the EngD programme.
Please note that not all places are advertised because sponsoring companies often have their own advertising processes for places.Read more
Respiratory Correlation Models in Image-Guided Radiotherapy
This 3 year EngD studentship is based in the Centre for Medical Image Computing (CMIC) and the Centre for Virtual Environments, Interaction and Visualisation (VEIV) at UCL, and sponsored by Elekta.
Respiratory motion causes errors and uncertainties when planning and delivering radiotherapy (RT) treatment to tumours in the thorax and abdomen, e.g. lung tumours and liver tumours. Current clinical practice is to either add margins to the target to account for the motion, thus irradiating more healthy tissue than for a static tumour, or to implant markers in the tumour so that its position can be determined during treatment, enabling gated delivery (when the RT bean is only switched on when the tumour is in the correct location) or tracked delivery (where the beam is made to follow the tumour’s motion).
On-board Cone-Beam CT systems are now commonly used to image a patient on the treatment couch just prior to the delivery of RT, to check that they are setup correctly and that the treatment is aligned with the tumour. Image acquisition times of approximately one minute mean that the reconstructed image cannot be used to determine the respiratory motion. However, the individual projection images, used to reconstruct the CBCT image, are acquired fast enough to capture the respiratory motion, but due to the nature of the images it is not possible to determine the anatomy and its motion from single images.
This project will build on previous work from CMIC on respiratory correlation models, which relate the internal motion of interest to an easily measured respiratory surrogate signal, such as the motion of the skin surface. The aim of this project will be to further develop these models so that they can be built from CBCT projection images, and to utilise these models for guiding RT delivery. This will involve close collaboration with researchers and engineers from Elekta as well as clinicians and RT physicists from the Institute of Cancer Research (ICR). The supervisors for the project are Dr Jamie McClelland and Prof David Hawkes at UCL, with co-supervisors from ICR and Elekta also advising on the project.