Abstract

Real-time magnetic resonance imaging (RT-MRI) is a safe and powerful tool for studying vocal tract dynamics during speech production. Emerging low-and mid-field MRI platforms bring new capability including higher frame rates, improved tissue contrast, and reduced blurring compared to conventional 1.5 T and 3T MRI. Here, we present a state-of-the-art speech production MRI protocol tailored to the 0.55 T platform. This includes 2D mid-sagittal vocal tract RT-MRI, simultaneous multislice RT-MRI with 3 parallel slices, tagged RT-MRI for functional evaluation of internal tongue deformation, biofeedback that allows participants to see their scan in real-time, and 3D static imaging. Imaging performance is comparable and, in several cases, superior to 1.5 T and 3T MRI, making 0.55 T an exciting new platform for speech research.

Metadata

publication

Proc. Interspeech 2024, 2590-2594, 2024

year

2024

publication date

2024

authors

Prakash Kumar, Ye Tian, Yongwan Lim, Sophia X Cui, Christina Hagedorn, Dani Byrd, Uttam K Sinha, Shrikanth Narayanan, Krishna S Nayak

link

https://www.isca-archive.org/interspeech_2024/kumar24b_interspeech.pdf

resource_link

https://www.isca-archive.org/interspeech_2024/kumar24b_interspeech.pdf

conference

Proc. Interspeech 2024

pages

2590-2594