Rapid quantitative susceptibility mapping of intracerebral hemorrhage

• Published in: Journal of magnetic resonance imaging 2020-03, Vol.51 (3), p.712-718
• Main Authors: De, Ashmita, Sun, Hongfu, Emery, Derek J., Butcher, Kenneth S., Wilman, Alan H.
• Format: Article
• Language: English
• Subjects: Aged; Aged, 80 and over; Cerebral Hemorrhage - diagnostic imaging; Correlation coefficient; Correlation coefficients; echo planar imaging (EPI); Female; Field strength; Hemorrhage; Humans; Image Interpretation, Computer-Assisted; intracerebral hemorrhage (ICH); Iron; Magnetic Resonance Imaging; Magnetic susceptibility; Male; Mapping; Measurement methods; Medical imaging; Middle Aged; Neuroimaging; Observers; Prospective Studies; quantitative susceptibility mapping (QSM); Reproducibility of Results; Shot; Statistical analysis; Statistical tests; susceptibility‐weighted imaging (SWI); Tracking
• ISSN: 1053-1807; 1522-2586
• DOI: 10.1002/jmri.26850

Background Quantitative susceptibility mapping (QSM) offers a means to track iron evolution in hemorrhage. However, standard QSM sequences have long acquisition times and are prone to motion artifact in hemorrhagic patients. Purpose To minimize motion artifact and acquisition time by performing rapid QSM in intracerebral hemorrhage (ICH) using single‐shot echo planar imaging (EPI). Study Type Prospective method evaluation. Population/Subjects Forty‐five hemorrhages were analyzed from 35 MRI exams obtained between February 2016 and March 2019 from 27 patients (14 male / 13 female, age: 71 ± 12 years) with confirmed primary ICH. Field Strength/Sequence 3T; susceptibility‐weighted imaging (SWI) with 4.54‐minute acquisition and 2D single‐shot gradient EPI with 0.45‐minute acquisition. Assessment Susceptibility maps were constructed from both methods. Measurement of ICH area and mean magnetic susceptibility were made manually by three independent observers. Motion artifacts were quantified using the magnitude signal ratio of artifact‐to‐brain tissue to classify into three categories: mild or no artifact, moderate artifact, or severe artifact. The cutoff for each category was determined by four observers. Statistical Tests Pearson's correlation coefficient and paired t‐test using α = 0.05 were used to compare results. Inter‐ and intraclass correlation was used to assess observer variability. Results Using 45 hemorrhages, the ICH regions measured on susceptibility maps obtained from EPI and SWI sequences had high correlation coefficients for area (R2 ≥ 0.97) and mean magnetic susceptibility (R2 ≥ 0.93) for all observers. The artifact‐to‐tissue ratio was significantly higher (P