Magnetic resonance imaging of lung cancer in the presence of respiratory motion: Dynamic keyhole and audio visual biofeedback

Abstract

Breath-to-breath variations in breathing can cause image artefacts. Day-to-day variations can cause a disagreement of position and volume between planning and treatment throughout radiotherapy procedures, requiring a larger treatment margin and longer treatment time. An advanced radiotherapy system requires: (1) a fast imaging technique for the compensation of breathing variations and/or (2) a respiratory motion management technique for the control of breathing variations. A novel MRI reconstruction method called “Dynamic keyhole” was proposed as a fast imaging technique. This thesis investigated (1) the concept of this method in terms of the improvement in temporal resolution with healthy volunteer MRI datasets and (2) the applicability of real-time lung tumour localization in terms of the accuracy of tumour motion and shape with lung cancer patient MRI datasets. The dynamic keyhole method achieved an increase in imaging frequency by up to a factor of five when compared with full k-space methods whilst achieving sub-millimetre tumour motion accuracy and preserving tumour shape within 98%. AV biofeedback respiratory guidance was used for healthy volunteers and lung cancer patients. This thesis investigated the impact of AV biofeedback on (1) intra- and inter-fraction lung tumour motion using cine-MRI, (2) inter-fraction lung tumour position and intra-fraction tumour volume using breath-hold MRI and (3) the improvement in image quality and the reduction in scan time using respiratory-gated MRI. AV biofeedback respiratory guidance improved intra- and inter-fraction tumour motion and position reproducibility, and intra-fraction tumour volume consistency. In addition, it was found to improve image quality and reduce scan time. The performance of the dynamic keyhole method and AV biofeedback respiratory guidance shown in this thesis illustrates potential advantages of real-time tumour imaging and tumour motion management in the course of lung cancer radiotherapy.