Optimization of Acoustic Noise for Single-Shot Echo-Planar Imaging by Varying Echo Spacing

Single-shot echoplanar imaging (EPI) sequence is a commonly-used readout scheme for functional magnetic resonance imaging (fMRI). It acquires signal in a short period of time with loud acoustic noise, which could cause discomfort for patients and even pose risk for sensitive populations, as well as confound auditory fMRI studies. Though a variety of attempts have been made toward quiet EPI scans, none has considered both the noise level and the timbre. In this study, we investigated the effect of varying echo spacing and modified gradient waveform on sound pressure level and noise spectral entropy. We then used genetic algorithm to optimize both sound pressure level and spectral entropy for single-shot EPI sequence by varying the duration of each readout unit with a sinusoidal waveform, changing the timbre significantly with increased entropy and reduced loudness. The resulting image quality were also compared with images obtained by standard EPI sequence.