A method for fast 3D imaging of contrast enhanced vessels or catheters using magnetic resonance projections. (September 2016)

• Record Type: Journal Article
• Authors: Unan, M.; Seimenis, I.; Webb, A.G.; Tsekos, N.V.
• Is Part Of: Physica medica. Volume 32(2016)Supplement 3
• Page Start: 248
• Publication Date: 2016-09
• Subjects: Medical physics -- Periodicals; Biophysics -- Periodicals; Biophysics -- Periodicals; Imagerie médicale -- Périodiques; Radiothérapie -- Périodiques; Rayons X -- Sécurité -- Mesures -- Périodiques; Physique -- Périodiques; Médecine -- Périodiques; 610.153
• DOI: 10.1016/j.ejmp.2016.07.527
• Languages: English
• ISSNs: 1120-1797

Abstract

Abstract : Introduction: With a plethora of soft-tissue contrast mechanisms, lack of ionizing radiation and on-the-fly computer controller adjustment of imaging parameters, MRI has emerged as an alternative modality for guiding interventions. However, due to its inherent low signal sensitivity, conventional MRI cannot achieve the high speeds of X-ray fluoroscopy. Purpose: To address this, we describe a novel approach for 3D MRI of tubular structures such as blood vessels or catheters, based on the collection of thick slab spatially matched projections. Materials and methods: The implemented method includes the following: three elements. (1) Collection of three orthogonal projections of the same volume that contains the structure with a GRE (TR/TE = 26.07/3.71 ms, angle = 75°, matrix = 256 × 256, FOV = 200 × 200 mm 2, slice = 200 mm). (2) Segmentation of the 2D structures on the three projections. (3) Reconstruction of the 3D structure by back projection. The method was tested on phantoms with vessel-mimicking structures made of tubing filled with 2% Gd-agent in a fatty matrix. The ground truth was an MRA (128 slices, TR/TE = 3.8/1.52 ms, angle = 40°, matrix = 384 × 264, FOV = 191 × 131 mm 2, slice = 1.3 mm). Results: The 3D centerline of the rendered structures was extracted and then found to be virtually the same (±pixel) to this extracted from a multislice MRA of the same structure. Conclusion: The method can accurately image 3D tubular object in 20 s as compared to 186 s with the used MRA. Disclosure: None of the authors has anything to declare.