Optimizing image guidance quality for high-accuracy proton therapy

Lead Research Organisation: University of Manchester
Department Name: School of Medical Sciences

Abstract

Background: Proton therapy is a type of advanced radiotherapy, with the potential of more precisely delivering a targeted dose of radiation to the tumour while improving sparing of surrounding healthy tissues. Hence, proton treatments can be as effective as conventional radiotherapy with fewer side effects. High-precision radiotherapy treatments require high-quality imaging at the treatment machine such as cone beam CT (CBCT). CBCT images allow changes in anatomy to be picked up and the treatment adjusted accordingly (a process called "adaptive radiotherapy"). Historically, one of the challenges for proton therapy is that its imaging technology lags behind that used in conventional radiotherapy. As a result, some challenging tumour sites (such as lung or head and neck cancers, where the tumours can move or shrink considerably during treatment) are rarely treated with proton therapy today. The proton therapy centre in Manchester is one of the first in the world to incorporate CBCT, and is in a unique position because of the presence of word-leading experts on the use of CBCT in radiotherapy in Manchester.

Objectives: As one of the first proton centres in the world to incorporate CBCT, we have the opportunity to be the first to evaluate current imaging techniques and optimize them given the characteristics and geometry of the proton therapy machine, aiming at setting new standards. We aim to improve the quality of images for proton therapy by developing specific image acquisition techniques, in particular dual-energy cone beam computed tomography (DE-CBCT), while optimising the imaging dose received by the patients. With improved image quality, we will implement adaptive treatment strategies for different tumour sites (e.g. lung cancer, head and neck cancer, paediatric patients). This will ensure that patient selection for proton therapy will be based on clinical factors (i.e. who benefits the most), instead of technical factors.

Methods: Dr Aznar's previous research has demonstrated that dual energy imaging, combining low- and high-energy x-ray images, can improve the contrast between the tumour and healthy tissues [2], as well as the definition of the tumour volume itself [3]. In addition, dual energy imaging enables a better characterisation of the daily "path" (or "range") of protons through the different tissues of the body [4]. However, the optimal image reconstruction will depend on the part of the body images (head vs thorax) or the organs that need to be visualised. Hence, different image combinations must be defined for different patient groups.
We will develop streamlined treatment verification and dose accumulation processes based on dual energy CBCT images and deformable image registration. In a first step, conventional CBCTs from a large database of head and neck and lung cancer patients treated at the Christie with standard photon radiotherapy will be used to obtain clinical relevant data about changes in anatomy/tumour volume during the course of treatment. Data from 1700 lung cancer patients has already been collected through work complementary to this project, including an extensive software infrastructure for automatic data analysis. Similar efforts are ongoing for other tumour sites. In a parallel effort, imaging phantoms will be used to investigate the best image acquisition parameters (considering contrast, signal to noise ratio as well as imaging dose to the patient) on the proton therapy machine, taking into accounts its unique characteristics, geometry and requirements compared to conventional radiotherapy applications. Prof van Herk's research has developed motion compensation techniques for CBCT, which are essential to develop DE-CBCT applications in mobile anatomy.

EPSRC areas: Medical imaging (including medical image and vision computing)
Healthcare technologies.

Publications

10 25 50

Studentship Projects

Project Reference Relationship Related To Start End Student Name
EP/N509565/1 01/10/2016 30/09/2021
2117618 Studentship EP/N509565/1 21/09/2018 30/06/2022 Josh Lindsay