Detecting Cerebral Microbleeds on Susceptibility-Weighted MR Imaging: Comparison with the Conventional MR Imaging Sequences

The purpose of our study was to compare the susceptibility-weighted imaging (SWI) with the conventional MR imaging sequences for their ability to detect cerebral microbleeds. We studied 17 patients (9 men and 8 women, mean age: 58 years) with microbleeds. The gradient-recalled echo T2*-weighted imaging (GRE), the T2-weighted imaging (T2WI), the fluid-attenuated inversion recovery (FLAIR) imaging and the SWI were obtained from all the subjects. The number of microbleeds was counted on each MR imaging sequence. We also compared the number of detected microbleeds (%) of each imaging technique with that of SWI being 100%. The Friedman ANOVA Test was used for statistical analysis and a p value of less than 0.05 was considered statistically significant. The number of detected mircobleeds on each imaging technique (%, mean ± standard deviation), as compared to that detected on SWI as 100%, was as follows: GRE vs. SWI (64.9 ± 19.3 vs. 100, p =0.001), T2WI vs. GRE (16.9 ± 17.7 vs. 64.9 ± 19.3, p = 0.001), and T2WI vs. FLAIR (16.9 ± 17.7 vs. 13.8 ± 17.7, p = 0.027). SWI demonstrates a significantly higher sensitivity for detecting microbleeds as compared to that of GRE and the other conventional MR imaging sequences