Abstract
PURPOSE: The aim of the present study is to develop a simple and practical method for measuring the signal-to-noise ratio (SNR) of magnetic resonance images called combined maximum b-value and echo time (COMBET) that could be suitable for pulse sequences to which a diffusion gradient can be applied. METHODS: In the COMBET method, we first obtain a signal image using the objective pulse sequence. Then, we obtain the noise image of this sequence using the diffusion gradient with the largest b-value and longest echo time. However, other imaging parameters are the same as those used for the signal image acquisition. The SNR is calculated from the mean signal intensity in the region of interest (ROI) of the signal image divided by the signal standard deviation in the ROI of the noise image after the required corrections. We compared SNRs determined using the COMBET and double echo with the longest second echo time (DELSET) methods for single-shot echo-planar imaging and fast spin-echo sequences in white mineral oil phantom, purified water phantom, human head, and upper abdomen. We used the subtraction method as the reference standard. RESULTS: The COMBET method could obtain the optimal noise image, whereas the DELSET method could not sufficiently suppress the long T(2) signal in the purified water phantom, cerebrospinal fluid, and digestive fluid. Therefore, the DELSET method afforded incorrect results for the long T(2) regions in the noise and SNR maps, while the COMBET method enabled the in vivo evaluation of the SNR even in the long T(2) regions. CONCLUSION: The COMBET method allows simple and practical SNR measurement, which is applicable to tissues with long T(2) relaxation time.