Room: AAPM ePoster Library
Purpose:
To evaluate the potential of utilizing image-based custom immobilization device as replacement without re-simulation should a posterior immobilization device made during simulation is misplaced, damaged or becomes unusable.
Methods:
One volunteer subject underwent pelvic MR-only simulation with standard alpha cradle immobilization. From the initial simulation images, a custom machine-milled Styrofoam posterior mold was fabricated. The volunteer returns for the same MR simulation with the custom image-based immobilization (IBI) device and the original alpha cradle for setup alignment comparison. Rigid registration with bony alignment was performed and the mutual information, root mean square difference and rotations were calculated. The external body contour of all three scans were created and the Dice similarity coefficient, Hausdorff distance (HD), mean distance to agreement (MDA), and Jaccard similarity index were used to perform pairwise comparisons between the external body contours from all three scans.
Results:
Evaluation of the rigid registration from the IBI scan and the alpha cradle scan to the initial alpha cradle scan showed that scanning with IBI resulted in superior mutual information, root mean square difference, and smaller rotation angles. The external contour comparison between the initial alpha cradle scan to the IBI scan achieve Dice similarity of 0.98, Jaccard index of 0.96, HD of 12.15 mm, and MDA of 2.18 mm. All metrics are comparable to the pairwise comparison between the external contour of the new alpha cradle scan and the initial alpha cradle scan.
Conclusion:
Custom image-based immobilization is an excellent alternative if the initially created immobilization device created during simulation is misplaced. The fabrication and use of such devices can eliminate the need to re-simulate and potentially re-plan.
Not Applicable / None Entered.
TH- RT Interfraction Motion Management: setup errors, immobilization, localization