Virtual Library
Start Your Search
M. Van Herk
Author of
-
+
P3.14 - Radiotherapy (ID 730)
- Event: WCLC 2017
- Type: Poster Session with Presenters Present
- Track: Radiotherapy
- Presentations: 1
- Moderators:
- Coordinates: 10/18/2017, 09:30 - 16:00, Exhibit Hall (Hall B + C)
-
+
P3.14-010 - Magnetic Resonance Imaging (MRI) for Lung Cancer Radiotherapy Planning and Treatment (ID 10085)
09:30 - 09:30 | Author(s): M. Van Herk
- Abstract
Background:
A potential benefit of integrating MR imaging in the radiotherapy planning and treatment process is the improved soft tissue contrast. However, this benefit may be obscured if there are substantial variations in the way that both target and organs at risk (OARs) are contoured, either at the time of initial simulation or at each fraction. The aim of this work is to develop standardised MR imaging sequences for target and OAR contouring in lung cancer patients for integration into the planning and treatment process for MR image-guided radiotherapy.
Method:
11 lung cancer patients were recruited across 3 European centres using a standardised MR protocol with various soft tissue contrasts as well as numerous respiratory management techniques. All MR images were acquired at 1.5 T with 3.5 mm slice thickness. 7 radiation oncologists and 3 radiologists reviewed the MR images alongside CT and FDG-PET-CT, rigidly registered to tumour area, to determine the most adequate sequences for OAR and tumour visualisation/delineation.
Result:
The most adequate MR sequences were found to be: Radial 3D T1, Turbo Spin Echo (TSE) fat-suppressed (fat-sat), DIXON TSE and T2 TSE (table 1). The sequences aided in visualisation of tumour, nodes and OARs including oesophagus, proximal bronchial tree, trachea, heart, lungs, spinal cord/canal and brachial plexus. The TSE (fat sat) and DIXON TSE gave strong signal from tumour, nodes and brachial plexus, aiding in their visualisation. The motion averaging characteristics ofthe Radial 3D T1 improve image consistency over slices, whereas the TSE sequences can include motion induced distortions.Table 1. The most adequate sequences identified for tumour, node and OAR visualisation. These sequences have been acquired with varying contrast and different respiratory management techniques.
Sequence Respiratory Management Purpose Coverage Radial 3D T1 (fat sat) Radial k-space sampling Tumour, nodes, OARs, spatial reference Ful thorax TSE (fat sat) Respiratory triggered Tumour, nodes, OARs Tumour (if tumour below aortic arch) DIXON TSE Multiple signal averages Brachial plexus, tumour, nodes Tumour and Brachial Plexus (if tumour above aortic arch) T2 TSE Respiratory triggered Tumour, nodes, OARs Full thorax
Conclusion:
A set of MR sequences suitable for lung cancer radiotherapy planning have been identified which provide adequate visualisation of tumour, nodes and OARs. The next step is to acquire these sequences in a set of patients and assess intra-/inter-clinician variability in target and OAR delineation, and develop an MR lung contouring atlas.