Application of intelligent imaging sensors to image guided and intensity modulated radiotherapy

Lead Research Organisation: University of Sheffield
Department Name: Electronic and Electrical Engineering

Abstract

Novel active pixel sensors have been developed under the Multidisciplinary Integrated Intelligent Imaging (Mi3) Project (funded by a Joint Research Council Basic Technology Award), which Sheffield leads and Institute of Cancer Research (ICR) are partners. This will lead to the realisation of radiation sensors that have increased functionality and integrated intelligence for use in a range of imaging applications within the field of medical imaging. This proposed research aims to exploit these latest developments in imaging and detector technology for application to two challenges in image guided radiotherapy (IGRT) and intensity modulated radiotherapy (IMRT), which will help the optimisation of conformality of dose delivery in cancer treatment with radiotherapy. The first challenge is to improve dose delivery accuracy to moving tumours in sites such as the lung. On of the most promising techniques for dealing with motion during treatment employs IGRT with X-ray fluoroscopy. Using two dedicated fluoroscopy units, tumour motion has been imaged continuously with the aid of X-ray radio-opaque markers. However, serious problems with this system exist: there is a significant latency period between the time at which the tumour is imaged and the identification of the tumour position and the cost of using such systems in terms of unwanted dose is non-trivial. If these systems are to be realised lower radiation intensity must be used which inevitably leads to poorer image quality and lower tracking accuracy. Lower tracking accuracy will lead to a greater number of errors or impractical treatment times. The second challenge is to optimise treatment verification, which due to the complex nature of IMRT delivery, is a time-consuming and therefore resource-heavy process. Groups, including the ICR Group, have pioneered the use of electronic portal imaging devices for IMRT verification. Imaging solutions for the verification of the position of the radiation beam during dynamic beam delivery with respect to the treatment plan have been presented, however, fully automated verification has yet to be implemented in day-to-day clinical practice.It is against this background that the project will address the question: given the image quality limitations of current portal imaging devices and the high dose rates imparted by fluoroscopic systems, is it feasible to increase tracking accuracy of intra-fractional tumour motion and the efficiency of IMRT verification using intelligent pixels sensors? Intelligent sensors will be built that have novel functionality that will potentially be useful for speeding up IMRT verification and optimising the tracking process. For example, these sensors will have region of interest read out and self-triggered detector read out that have advantages for increased speed of acquisition and data reduction. These functions can be used to make better use of the radiation available allowing us to acquire the minimum amount of data required to make decisions regarding tumour marker position or field leaf position for IMRT verification. A prototype imaging system will be designed, constructed and tested in order to investigate data acquisition methods that can be implemented with active pixel sensors to fully optimise image acquisition and hence make portal/fluoroscopic imaging a viable solution for IGRT and efficient verification. Optimisation of tracking methods through the use of intelligent sensors will be carried out firstly through the development of tracking algorithms that can be implemented within the image sensor, through simulation of this hard ware processes and through testing using the novel sensors and field programmable gate arrays. The final stage of this project will include the development of a concept demonstrator for intelligent sensor based radiotherapy imaging.

Publications

10 25 50
publication icon
Konstantinidis A (2010) Optical characterisation of a CMOS active pixel sensor using periodic noise reduction techniques in Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment

Related Projects

Project Reference Relationship Related To Start End Award Value
EP/F038518/1 14/11/2008 01/01/2011 £260,386
EP/F038518/2 Transfer EP/F038518/1 01/01/2011 31/10/2011 £86,189