Fast Chemical Shift Encoded and J-Decoupled/J-Resolved MRSI Based on Cross-Term Spatiotemporal Encoding

Magnetic resonance spectroscopic imaging (MRSI) is a crucial noninvasive technique for spatially resolving variation in brain metabolites, aiding the diagnosis of brain diseases and nervous system pathologies. Traditional MRSI approaches have significant challenges because of the long acquisition time, limiting their practical applications. Echo-planar spectroscopic imaging (EPSI) has brought revolutionary improvement in the time efficiency of data collection by simultaneously encoding one spatial and the temporal dimension using echo-planar fast gradient oscillation during signal readout. However, the use of rapid oscillating readout gradients in EPSI constrains its spectral bandwidth and increases sensitivity to field inhomogeneities. Cross-term spatiotemporal encoding (xSPEN), which relies on rapid gradient oscillations, utilizes chirp pulses to implement spatiotemporal encoding, providing a single-shot MRI technique with superior resistance to chemical shifts and field inhomogeneities. In this study, we extended xSPEN to perform echo-planar imaging-based acquisition with increasing time ( $t_1$ ) evolution, which we term $t_1$ -xSPEN spectroscopic imaging. This approach enhances sampling efficiency and allows for flexible spectral bandwidth of MRSI, surpassing the constraints of EPSI. By splitting the $t_1$ evolution into $\tau -t_1/2$ and $t_1/2$ parts on either side of a $\pi$ pulse, $t_1$ -xSPEN achieves a constant- $\tau$ J-coupling, enabling J-decoupled xSPEN spectroscopic imaging. Furthermore, integrating this approach with turbo spin echo train evolution enables robust and relatively distortion-free acquisition of J-coupling information. This novel method holds promise for improving the resolution and accuracy of metabolite mapping in the brain, offering new insights into the diagnosis and understanding of neurological disorders.