Premium
3D imaging using magnetic resonance tomosynthesis (MRT) technique
Author(s) -
Kim MinOh,
Zho SangYoung,
Kim DongHyun
Publication year - 2012
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 - 10.1118/1.4737111
Subject(s) - tomosynthesis , imaging phantom , iterative reconstruction , projection (relational algebra) , magnetic resonance imaging , image resolution , tomographic reconstruction , medical imaging , computer science , tomography , computer vision , artificial intelligence , optics , physics , mammography , algorithm , radiology , medicine , cancer , breast cancer
Purpose: To introduce an alternative approach to three‐dimensional (3D) magnetic resonance (MR) imaging using a method that is similar to x‐ray tomosynthesis. Methods: Variable angle tilted‐projection images are acquired using a multiple‐oblique view (MOV) pulse sequence. Reconstruction is performed using three methods similar to that of x‐ray tomosynthesis, which generate a set of tomographic images with multiple 2D projection images. The reconstruction algorithm is further modified to reformat to the practical imaging situations of MR. The procedure is therefore termed magnetic resonance tomosynthesis (MRT). To analyze the characteristics of MRT, simulations are performed. Phantom and in vivo experiments were done to suggest potential applications. Results: Simulation results show anisotropic features that are structurally dependent in terms of resolution. Partial blurrings along slice direction were observed. In phantom and in vivo experiments, the reconstruction performance is particularly noticeable in the low SNR case where improved images with lower noise are obtained. Reformatted reconstruction using thinner slice thickness and/or extended field‐of‐view can increase spatial resolution partially and alleviate slice profile imperfection. Conclusions: Results demonstrate that MRT can generate adequate 3D images using the MOV images. Various reconstruction methods in tomosynthesis were readily adapted, while allowing other tomosynthesis reconstruction algorithms to be incorporated. A reformatted reconstruction process was incorporated for applications relevant to MR imaging.