Premium
TH‐A‐BRF‐12: Assessment of 4D‐MRI for Robust Motion and Volume Characterization
Author(s) -
GlideHurst C,
Kim J,
Hu Y,
Wen N,
Mutic S,
Chetty I
Publication year - 2014
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.4889558
Subject(s) - nuclear medicine , coronal plane , duty cycle , imaging phantom , medicine , voxel , magnetic resonance imaging , physics , radiology , power (physics) , quantum mechanics
Purpose: Precise radiation therapy for abdominal lesions is complicated by respiratory motion and poor soft tissue contrast from 4DCT whereas 4DMRI provides superior tissue discrimination. We evaluated a novel 4D‐MRI algorithm for MR‐SIM motion management. Methods: Respiratory‐triggered, T2‐weighted single‐shot Turbo Spin Echo 4D‐MRI was evaluated for open high‐field 1.0T MR‐SIM. A programmable platform pulled objects on a trolley ∼2 cm superior‐inferior (S‐I) for “regular” (sinusoidal, (1‐cos 2 ), 3‐5 second periods) and “irregular” breathing patterns (exaggerated (1‐cos 2 ) and patient curves), while a respiratory waveform was generated via a pressure sensor device. Coronal 4D‐MRIs (2–12;10 phases, TE/TR/α = 35−61/6100 ms/90°, voxel=1×1×4 mm 3 ) were acquired for 54 unique phantom cases. Abdominal 4D−MRIs were evaluated for 5 healthy volunteers and 1 liver cancer patient (6–10 phases, TE/TR/α = 30−96/4500−6100 ms/90°, voxel=1×1×5–10 mm 3 ) on an IRB‐approved protocol. Duty cycle, scan time, and excursion were evaluated between phase acquisitions and compared to coronal cine‐MRI (∼1 frame/sec). Maximum intensity projections (MIPs) were analyzed. Results: In phantom, average duty cycle was 42.6 ± 11.4% (range: 23.6–69.1%). Regular, periodic breathing (sinusoidal, (1‐cos 2 )) yielded higher duty cycles than irregular (48.5% and 35.9%, respectively, p<0.001) and fast periods had higher duty cycles than slow (50.4% for 3s and 39.4% for 5s, p<0.001). ∼4‐fold acquisition time increase was measured for 10‐phase versus 2‐phase. MIP renderings revealed that SI object extent was underestimated a maximum of 4% (3mm) and 8% (6mm) for cine and 2‐phase 4D‐MRI, respectively, with respect to 10‐phases. However, this was waveform dependent. A highly irregular breathing volunteer yielded lowest duty cycle (23%) and longest 10‐phase scan time (∼14 minutes), although 6‐phase acquisition for a liver cancer patient was reasonable (50% and 7.4 minutes, respectively). Conclusion: 4D‐MRI offers potential for excursion characterization, although results suggest the use of adequate phases is important. Further application and clinical validation are warranted. Research supported in part by a grant from Philips HealthCare (Best, Netherlands) and an Internal Mentored Grant from Henry Ford Health System.