Sequence-based acoustic noise reduction of clinical MRI scans

Abstract

Purpose: Clinical MRI patients typically experience elevated acoustic noise levels of 80-110 dB(A). In this study, standard clinical turbo spin echo (TSE) and gradient echo (GRE) sequences were optimized for reduced acoustic noise at preserved diagnostic image quality. Methods: The physical sources of acoustic noise generation in an MRI gradient coil were analyzed. A sequence conversion algorithm was derived that optimized the gradient time scheme for an arbitrary MRI sequence, preserving the governing spin physics. The algorithm was applied to generate "quiet" versions of standard clinical TSE and GRE sequences. Results: The first volunteer images indicated that contrast-to-noise ratio and perceived diagnostic image quality remained on the same level for the algorithmic optimization. Additional careful TSE- and GRE-specific protocol adaptions yielded total acoustic noise reductions of up to 14.4 dB(A) for the TSE and up to 16.8 dB(A) for the GRE. Conclusion: A physical sound pressure reduction of 81% (TSE) and 86% (GRE) for MRI patients was achieved. The results can be used to render MRI scans more patient-friendly in clinical practice, particularly for patients who are young, scared, or elderly.