Improved Simultaneous Multislice EPI Reconstruction for Diffusion MRI using regularized iterative phase...

Simultaneous multi-slice echo planar imaging (SMS-EPI) has significantly advanced diffusion MRI by enabling efficient whole-brain coverage at high spatial resolution. Despite the success of existing phase error-corrected SENSE (PEC-SENSE) in eliminating Nyquist ghosts, noise amplification and residual artifacts pose substantial challenges at high acceleration factors. We propose an iterative reconstruction solution that integrates phase error-corrected SENSE (iPEC-SENSE) with wavelet regularization. This method addresses the noise amplification and artifact issues found in existing PEC-SENSE reconstructions. The regularized iPEC-SENSE method demonstrated robust suppression of artifacts and noise, significantly improving image quality in diffusion MRI with SMS-EPI. This enhanced reconstruction quality facilitates more reliable diffusion tensor imaging (DTI) analysis in terms of reduced coefficient of variation and test-retest variation. The proposed iPEC-SENSE approach effectively mitigates the challenges of noise amplification in existing PEC-SENSE for SMS-EPI, while maintaining the ability to eliminate Nyquist ghosts, providing a more precise and reproducible solution for diffusion MRI. This work offers a robust parallel imaging reconstruction method for improving the quality and accuracy of diffusion MRI with SMS-EPI, enhancing the potential for advanced neuroimaging applications.