Published in:
01-04-2018 | MAGNETIC RESONANCE IMAGING
Short tau inversion recovery in breast diffusion-weighted imaging: signal-to-noise ratio and apparent diffusion coefficients using a breast phantom in comparison with spectral attenuated inversion recovery
Authors:
Tsukasa Yoshida, Atsushi Urikura, Kensei Shirata, Yoshihiro Nakaya, Masahiro Endo, Shingo Terashima, Yoichiro Hosokawa
Published in:
La radiologia medica
|
Issue 4/2018
Login to get access
Abstract
Objective
This study aimed to compare the signal-to-noise ratios (SNRs) and apparent diffusion coefficients (ADCs) obtained using two fat suppression techniques in breast diffusion-weighted imaging (DWI) of a phantom.
Materials and methods
The breast phantom comprised agar gels with four different concentrations of granulated sugar (samples 1, 2, 3, and 4). DWI with short tau inversion recovery (STIR-DWI) and that with spectral attenuated inversion recovery (SPAIR-DWI) were performed using 3.0-T magnetic resonance imaging, and the obtained SNRs and ADCs were compared. ADCs were also compared between the right and left breast phantoms.
Results
For samples 3 and 4, SNRs obtained using STIR-DWI were lower than those obtained using SPAIR-DWI. For samples 2, 3, and 4, overall ADCs obtained using STIR-DWI were significantly higher than those obtained using SPAIR-DWI (p < 0.001 for all), although no significant difference was observed for sample 1 (p = 0.62). STIR-DWI shows a positive bias and wide limits of agreement in Bland–Altman plot. The coefficients of variance of overall ADCs were good in STIR-DWI and SPAIR-DWI. For all samples, STIR-DWI demonstrated slightly larger percentage differences in ADCs between the right and left phantoms than SPAIR-DWI.
Conclusion
SNRs and ADCs obtained using STIR-DWI are influenced by the T
1 value; a shorter T
1 value decreases SNRs, overestimates ADCs, and induces the measurement error in ADCs. STIR-DWI showed a larger difference in ADCs between the right and left phantoms than SPAIR-DWI.