Room: Exhibit Hall | Forum 2
Purpose: After a success in developing super-resolution (SR) time-resolved (TR) 4DMRI, we plan to extend the SR approach to reconstruct T2w TR-4DMRI with higher soft-tissue contrast using deformable image registration (DIR) to facilitate clinical applications for multi-breath motion assessment and tumor delineation.
Methods: The SR-based T1w TR-4DMRI was reconstructed by deforming a high-resolution (2x2x2mm³) breath-hold (BH) 3DMRI image to a series of low-resolution (5x5x5mm³) 3D cine images at a 2Hz frame rate in free breathing (FB). The deformation vector fields (DVFᵀ¹ʷ�ᴰ) was saved. To achieve SR-based T2w TR-4DMRI, we developed two approaches. In the first method, a high-resolution (2x2x2mm³) T2w 3DMRI image was acquired, deformed to the T1w BH image, and this DVFᵀ²ʷᴴᴿ was combined with DVFᵀ¹ʷ�ᴰ to reconstruct T2w TR-4DMRI. In the second method, the DIR was reversed with DVFᵀ¹ʷᴴᴿ (=[DVFᵀ²ʷᴴᴿ]�¹), the deformed T1w BH was used to reconstruct the T1w TR-4DMRI, and the new combined DVFs were used to reconstruct T2w TR-4DMRI. Two lung cancer patients from an IRB-approved protocol were selected for this feasibility study. To acquire the high-resolution 3DMRI image, we applied a navigator-triggered respiratory-correlated 4DMRI scanning protocol to acquire 3-bin images (full-exhalation, mid-inhalation, and full-inhalation) with minimal binning artifacts. The closest T2w 3DMRI to a T1w BH was used with minimal deformation. The tumor alignment among the 3D cine, T1w and T2w TR-4DMRI was compared.
Results: Excellent tumor alignment is achieved for T2w TR-4DMRI within 2mm, like T1w TR-4DMRI. The first method is convenient since the existing DVFT1w4D can be borrowed from T1w TR-4DMRI reconstruction, while the second method has the advantage of keeping the T2w tumor volume.
Conclusion: The feasibility of T2w TR-4DMRI reconstruction has been demonstrated. The image alignment in low soft-tissue contrast is adequate. A further study with more patients is ongoing to fully characterize T2w TR-4DMRI image quality for clinical applications.
Funding Support, Disclosures, and Conflict of Interest: The Memorial Sloan Kettering cancer center has a master research agreement with Philips Healthcare