4D Composite MR Image Distortion Field: Quantification and Applications for MRI-Guided Radiotherapy
T Stanescu*, T Tadic, D Jaffray, Princess Margaret Cancer Centre, Toronto, ONWE-G-WAB-3 Wednesday 4:30PM - 6:00PM Room: Wabash Ballroom
Purpose: To study the interplay effects of MR image distortion fields for moving/deformable anatomy and their implications for MR-guided radiotherapy.
Methods: The composite distortion field is given by the combination of two independent components a) the system-related distortions due to B0 inhomogeneities and imaging gradients nonlinearities and b) the tissue magnetic susceptibility-induced effects. Field a) is static while b) is dynamic, modifying as a function of tissue interface characteristics (shape, location, type of neighboring tissues). The field experienced by a moving or deformable (over time) target/organ can be very complex as entity changes location and shape. Field a) was measured with a linearity object in conjunction with a harmonic analysis. Field b) was computed numerically using a finite difference method. Both fields were quantified for moving/deformable anatomy using multiple image data bins derived from 4D CT and 4D MRI for several lung and liver cases.
Results: The 4D composite field tools and methodology was validated in phantoms. The independent and combined distortion fields were quantified for several lung and liver patients. For a lung case with a target travelling 2.4 cm in the superior-inferior direction, the susceptibility distortions for the inhale and exhale phases were 3.84 and 2.5 ppm, respectively. The system distortion difference for the two phases was about 1 mm. The maximum 4D composite distortion between the inhale and exhale was approximately 1.5 mm. For liver, the target experienced negligible distortions (embedded in soft-tissue) and the diaphragm showed a local distortion larger than 1 mm.
Conclusion: The combined effect due to various types of MR image distortions were quantified for organ motion/deformation. Geometric distortion introduced by the 4D composite field may be significant for certain scenarios. However, this can be mitigated by a careful selection of imaging and by implementing robust methods for image correction.