X Wu*, y zhang, H An, H Gach, S Goddu, H Li, D Yang, Washington University in St. Louis, St. Louis, MO

• 
  X Wu

• 
  

SU-D-TRACK 4-1 (Sunday, 7/25/2021) 2:00 PM - 3:00 PM [Eastern Time (GMT-4)]

Purpose: To treat tachycardia disease using the noninvasive cardiac radiation ablation, an accurate patient-specific heart motion model is required for treatment planning and motion management. In this study, we developed a procedure to generate a subject-specific 5D (3D+cardiac+respiration) heart motion model from free-breathing continuous 2D CINE MRI movies without additional ECG and respiratory signals.

Methods: Seventeen 2D CINE MRI movies of a healthy volunteer was acquired continuously in the sagittal view at 17 positions at 5mm intervals, covering the entire heart. Each CINE movie was acquired for 60 seconds (8f/s, 100x100x480, 1.5x1.5x5mm3). In each movie frame. The cardiac signal was extracted from the heart area, and the respiratory signal was extracted from the average image intensity of the selected box in the diaphragm. Each signal was fitted to a combination of short cine curves. The final cardiac and respiratory phase values were assigned based on the fitting results. 3D volumes of selected cardiac and respiratory phase pairs were constructed from the 2D frames at all sagittal positions based on the assigned values and the temporal and spatial continuity. The deformation fields among the constructed 3D volumes of the selected cardiac and respiratory phases were computed using a group-wise deformable registration method. A PCA analysis was applied on the computed deformation fields to create the final 5D motion models.

Results: The proposed procedure was applied for 2D cine MRIs of a healthy volunteer. The 5D motion model was composed successfully and 3D heart volumes at any cardiac and respiratory phases can be constructed using this model.

Conclusion: We developed a procedure to generate a 5D subject-specific heart motion model from continuous 2D CINE MRI movies. It will be useful in managing cardiac and respiratory motions in radiation therapy for cardiac diseases.

Funding Support, Disclosures, and Conflict of Interest: The project described was partially supported by the Agency for Healthcare Research and Quality (AHRQ) grant number R01-HS022888, National Institute of Biomedical Imaging and Bioengineering (NIBIB) grant R03-EB028427 and National Heart, Lung, and Blood Institute (NHLBI) grant R01-HL148210.

Handouts

Keywords

Signal Processing, 3D, Heart

Taxonomy

IM/TH- MRI in Radiation Therapy: General (most aspects)

Contact Email