Subscribe to RSS
DOI: 10.1055/s-2006-931854
DTI Measurements of Fractional Anisotropy at 3 Tesla: Sequence optimization and reproducibility
Purpose: To determine suitable parameters for rapid DTI acquisition at 3 tesla, to assess the reproducibility of fa-measurements which are one of the core information for fiber trackings, and to quantify the influence of noise reduction and averaging high b-values on resulting FA-maps.
Methods: The study collective consisted of 10 volunteers for the issue of sequence optimization and 2 volunteers for reproducibility assessment and testing for the use of noise decorrelation and high b averaging. Concerning sequence optimization, the following constant parameters were used: a single shot SENSE imaging protocol with 2×2x2mm spatial resolution, covering sixty slices and measuring 15 diffusion directions. Parameters varied during sequence optimization were b-values (400, 600, 800s/mm2) and SENSE-factors (1 to 4), with the number of trackable callosal fibers as the criterion for parameter comparison. To assess the reproducibility of DTI sequences, we performed three independent measurements per session on two separate days, and these sequences were acquired with and without averaging of high b-values and noise reduction, which resulted in a total of 12 acquisitions per day.
Results: The highest number of trackable callosal fibers, was achievable with a b-value of 600s/mm2 and a SENSE factor of 2. Acquisition times were ranking from a maximum of 10 down to 4 and a half minutes. Concerning reproducibility, the highest relative standard deviation around the mean of repeated acquisitions, was ranking around 6 percent. The impact of noise reduction algorithms and averaging of the high b-values was low, so that the use of noise decorrelation and averaging high b may be withdrawn in favour of acquisition time.
Conclusion: SENSE-DTI at 3 Tesla allows for rapid DTI acquisition and yields highly reproducible FA measurements.