Premium
Comparison of Magnetic Resonance Angiography Scans on 1.5, 3, and 7 Tesla Units: A Quantitative Study of 3‐Dimensional Cerebrovasculature
Author(s) -
Nowinski Wieslaw L.,
Puspitasaari Fiftarina,
Volkau Ihar,
Marchenko Yevgen,
Knopp Michael V.
Publication year - 2013
Publication title -
journal of neuroimaging
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.822
H-Index - 64
eISSN - 1552-6569
pISSN - 1051-2284
DOI - 10.1111/j.1552-6569.2011.00597.x
Subject(s) - volume (thermodynamics) , medicine , magnetic resonance angiography , magnetic resonance imaging , angiography , nuclear medicine , radiology , physics , quantum mechanics
BACKGROUND Although multiple studies demonstrate benefits of high field imaging of cerebrovasculature, a detailed quantitative analysis of complete cerebrovascular system is unavailable. To compare quality of MR angiography (MRA) acquisitions at various field strengths, we used 3‐dimensional (3D) geometric cerebrovascular models extracted from 1.5T/3T/7T scans.METHODS The 3D cerebrovascular models were compared in volume, length, and number of branches. A relationship between the vascular length and volume was statistically derived. Acquisition performance was benchmarked against the maximum volume at infinitive length.RESULTS The numbers of vessels discernible on 1.5T/3T/7T are 138/363/907. 3T shows 3.3(1.9) and 7T 1.2(9.1) times more arteries (veins) than 1.5T. The vascular lengths and volumes at 1.5T/3T/7T are 3.7/12.5/22.7 m and 15.8/26.6/28.0 cm 3 . For arteries: 3T‐1.5T gain is very high in length, high in volume; 7T‐3T gain is medium in length, small in volume. For veins: 3T‐1.5T gain is moderate in length, high in volume; 7T‐3T gain is very high in length, moderate in volume. 1.5T shows merely half of vascular volume. At 3T 6%, while at 7T only 1% of vascular volume is missing.CONCLUSION Our approach differs from standard approaches based on visual assessment and signal (contrast)‐to‐noise ratio. It also measures absolute acquisition performance, provides a unique length‐volume relationship, and predicts length/volume for intermediate teslages.