Encrypted login | home

Program Information

Can MR Image Acquisition with Stereotactic Head Frame Improve Image Fusion Consistency in SRS Treatment Planning?

no image available

Y DING1*, J Yuan1 , K Cheung1 , S Yu1 , (1) Hong Kong Sanatorium & Hospital, Hong Kong, Hong Kong,


MO-RPM-GePD-J(A)-6 (Monday, July 31, 2017) 3:45 PM - 4:15 PM Room: Joint Imaging-Therapy ePoster Lounge - A

Purpose: To study quantitatively if MR imaging of treatment object with stereotactic head frame can improve CT-MR image registration consistency in SRS treatment planning.

Methods: A coconut phantom with 4 PinPoint® markers attached on A/P and L/R surface was fixed on a Leksell® Coordinate Frame. MR images were acquired with the frame in a 1.5T MR simulator using 3D TurboFLASH with distortion correction. To simulate positional errors, three phantom positions were designed, where the second position included a longitudinal shift from the first position, and the third scan included a rotation from the second position. A set of CT image of the phantom was also acquired using CT simulator. CT-MR fusion for each position was performed in treatment planning workstation (Elekta Monaco). For frame-based technique, manual fusion and semi-automatic point registration fusion were employed based on head frame. For frameless technique, manual fusion and automatic fusion were employed regardless of head frame. PinPoint marker centroid locations were extracted from the registered MR images.

Results: The standard deviations of displacements among PinPoint markers were smaller in frame-based (0.2 and 0.8 mm for manual and semi-automatic fusion respectively) than frameless (1.0 and 0.9 mm for manual and automatic fusion respectively) techniques. These implied that fiducial landmark can potentially improve fusion consistency. Noticeable increases (up to 2.6 mm) in mean displacements between translation only and translation+rotation suggested that fusion accuracy might be vulnerable to the inclusion of rotational positional error, while this error could be minimized in frame-based setup due to fixation of the head frame. Smaller mean displacements were found in using automatic registration compared to manual registrations, which was consistent with that visual-based manual fusion was more subjective.

Conclusion: Our results suggest that stereotactic coordinate frame in MR scan can improve fusion consistency by reducing intra-observer variation and positioning variation.

Contact Email: