¹H-MRSI at 7T using spectral-selective binomial excitation and adiabatic refocusing

Li, Xinyu

Permalink

https://hdl.handle.net/2142/132804 Copy

Description

Title

¹H-MRSI at 7T using spectral-selective binomial excitation and adiabatic refocusing

Author(s)

Li, Xinyu

Issue Date

2025-12-08

Director of Research (if dissertation) or Advisor (if thesis)

Lam, Fan

Department of Study

Bioengineering

Discipline

Bioengineering

Degree Granting Institution

University of Illinois Urbana-Champaign

Degree Name

M.S.

Degree Level

Thesis

Keyword(s)

• Magnetic resonance spectroscopic imaging (MRSI)
• Ultra-high field
• Spin-Echo Echo-Planar Spectroscopic Imaging (SE-EPSI)

Abstract

Magnetic resonance spectroscopic imaging (MRSI) is a multiscale imaging modality that can noninvasively map metabolic changes in vivo. These metabolite maps have broad potential in clinical and preclinical applications, such as tumor detection and grading, and the study of metabolic alterations in Alzheimer’s disease and other neurological disorders. However, MRSI signals typically suffer from low signal-to-noise ratio (SNR) due to the inherently low sensitivity of MRSI. Ultra-high field systems such as 7T can provide higher SNR and better spectral separation of metabolite peaks, but they also introduce pronounced B₀ and B₁ field inhomogeneity, larger chemical shift displacement error (CSDE), stricter specific absorption rate (SAR) constraints, longer acquisition time, and technical challenges for implementing conventional spectroscopic sequences. In recent years, several novel RF pulses have been developed to address these challenges, either by reducing RF power deposition and CSDE or by implementing advanced k-space sampling patterns with model-based image reconstruction methods to achieve fast, high-resolution ¹H MRSI. However, most of these approaches are based on FID-MRSI, which provides higher SNR and lower SAR with short echo times (TEs) but limits the development of multi-TE ¹H MRSI, J-resolved MRSI, diffusion weighted imaging (DWI), and spectral editing (e.g., targeting short-T₂ components). This dissertation focuses on implementing these advanced RF pulses within a spin-echo echo-planar spectroscopic imaging (SE-EPSI) framework to improve the robustness and acquisition efficiency of ¹H MRSI. Prior work has demonstrated that gradient offset independent adiabaticity (GOIA) pulses can reduce SAR, provide improved B₁ insensitivity, and yield more uniform refocusing compared with conventional adiabatic pulses. On the other hand, binomial pulses offer strong spectral selectivity, B₁ insensitivity, and efficient water suppression. Motivated by these advantages, we combine these pulse designs with a union-of-subspace (UoSS) model and SPectroscopic Imaging by exploiting spatiospectral CorrElation (SPICE) to address the limitations of ultra-high-field metabolic mapping in both data acquisition and data processing. Specifically, we implemented: (1) A 1-3 ̅-3-1 ̅ binomial excitation pulse to achieve B₁-insensitive, spectrally selective excitation that does not require an additional water-suppression module, and (2) A pair of GOIA refocusing pulses to reduce SAR, produce a uniform slice profile, and further improve insensitivity to B₁ inhomogeneity. To evaluate and validate the new sequence, we performed numerical simulations and in vitro and in vivo experiments using a water tube and a standard spectroscopic phantom. The proposed binomial excitation achieved a water-signal reduction by a factor of approximately 500. The RF power of the GOIA adiabatic refocusing pulse was reduced to about half that of a conventional hyperbolic secant refocusing pulse, while its effective bandwidth was increased by a factor of 10. We expect that the new MRSI sequence, combining a spectrally selective 1-3 ̅-3-1 ̅ binomial excitation pulse with spatial-selective GOIA refocusing pulses, will help overcome key limitations at ultra-high field and enable more reliable metabolic mapping in in vivo studies. The proposed sequence demonstrates strong potential for high-resolution metabolic imaging in future clinical applications.

Graduation Semester

2025-12

Type of Resource

Thesis

Handle URL

https://hdl.handle.net/2142/132804

Copyright and License Information

Copyright 2025 Xinyu Li

Owning Collections