High b‐Value and Ultra‐High b‐Value Diffusion Weighted MRI in Stroke

• Published in: Journal of magnetic resonance imaging Vol. 61; no. 3; pp. 1459 - 1469
• Main Authors: Jiang, Rifeng, Wang, ZhenXiong, Liu, Jun, Li, Ting, Lv, YanChun, Xie, Chuanmiao, Su, Changliang
• Format: Journal Article
• Language: English
• Published: Hoboken, USA John Wiley & Sons, Inc 01.03.2025; Wiley Subscription Services, Inc
• Subjects: Adult; Aged; Aged, 80 and over; Anisotropy; Brain - diagnostic imaging; Brain Ischemia - diagnostic imaging; Correlation analysis; Diffusion coefficient; Diffusion Magnetic Resonance Imaging - methods; diffusion weighted MR imaging; Female; Field strength; Humans; Image Interpretation, Computer-Assisted - methods; Image Processing, Computer-Assisted - methods; Ischemia; ischemic infarction; Lesions; Male; Masks; Middle Aged; Noninvasive evaluation; Prospective Studies; Rank tests; Reproducibility of Results; Statistical analysis; Statistical tests; Stroke - diagnostic imaging; Substantia alba; White Matter - diagnostic imaging; diffusion weighted MR imaging; ischemic infarction
• ISSN: 1053-1807; 1522-2586; 1522-2586
• DOI: 10.1002/jmri.29547

Purpose To explore the application value of high‐b‐value and ultra‐high b‐value DWI in noninvasive evaluation of ischemic infarctions. Study Type Prospective. Subjects Sixty‐four patients with clinically diagnosed ischemic lesions based on symptoms and DWI. Field Strength/Sequence 3.0 T/T2‐weighted fast spin‐echo, fluid‐attenuated inversion recovery, pre‐contrast T1‐weighted magnetization prepared rapid gradient echo sequence, multi‐b‐value trace DWI and q‐space sampling sequences. Assessment Lesions were segmented on standard b‐value DWI (SB‐DWI, 1000 s/mm2), high b‐value DWI (HB‐DWI, 4000 s/mm2) and ultra‐high b‐value DWI (UB‐DWI, 10,000 s/mm2), and cumulative segmented areas were the final abnormality volumes. Normal white matter (WM) areas were obtained after binarization of segmented brain. In 47 patients, fractional anisotropy (FA) and apparent diffusion coefficients (ADCs) at b values of 1000, 4000, and 10,000 s/mm2 were extracted from symmetrical WM masks and lesion masks of contralateral WM (CWM) and lesion‐side WM (LWM). Statistical Tests Wilcoxon matched‐pairs signed‐rank test and Pearson correlation analysis. Two‐tailed P‐values <0.05 were considered statistically significant. Results Various signals of HB‐/UB‐DWI (hypo‐, iso‐ or hyper‐intensity) were observed in strokes compared with SB‐DWI, and some areas with iso‐intensity of SB‐DWI manifested with hyper‐intensity on HB‐/UB‐DWI. Abnormality volumes from SB‐DWI were significantly smaller than those from HB‐DWI and UB‐DWI (10.32 ± 16.45 cm3, vs. 12.25 ± 19.71 cm3 and 11.83 ± 19.41 cm3), while no significant difference exist in volume between HB‐DWI and UB‐DWI (P = 0.32). In CWM, FA significantly correlated with ADC4000 and ADC10,000 (maximum r = −0.51 and −0.64), but did not significantly correlate with ADC1000 (maximum r = −0.20, P = 0.17). ADC1000 or ADC4000 of LWM not significant correlated with FA of CWM (maximum r = −0.28, P = 0.06), while ADC10,000 of LWM significantly correlated with FA of CWM (maximum r = −0.46). Data Conclusion HB‐ and UB‐DWI have potential to be supplementary tools for the noninvasive evaluation of stroke lesions in clinics. Evidence Level 2 Technical Efficacy Stage 2