Optimization of Magnetic Resonance Spin Tomography in Time-domain sequences

Publication date

2025-10-28

Authors

Fuderer, MihaORCID 0000-0002-5673-915X

Editors

Advisors

Supervisors

van den Berg, Cornelis A TORCID 0000-0002-5565-6889
Sbrizzi, AlessandroORCID 0000-0003-3276-4542ISNI 0000000396833383

Document Type

Dissertation

Collections

Open Access logo

License

Abstract

MR-STAT is an efficient MRI technique to simultaneously image several quantitative parameters of the human body. “Quantitative”, as opposed to the current practice of “weighted” images, allows for comparison of results over different hospitals. MR-STAT is a “transient-state” technique and relies heavily on applying time-varying radio-frequency (RF) pulses. Yet, some time-variation patterns can yield substantially better results than other patterns. Here, “better” primarily refers to higher precision within a given scan time. In this work, a tool is presented (“BLAKJac”) that searches for optimized RF pulse patterns. Importantly, it is shown that it is very beneficial to not only vary the amplitude of the RF-pulses, but to also optimally vary their phase. Although aforementioned optimization focuses on precision, two important causes of inaccuracy are studied in detail: the influence of water diffusion and the influence of an imperfect spatial transmit field pattern (“the B1+ field”). The effects of these imperfections also strongly depend on the chosen RF-pulse pattern. The primary benefit of quantitative MR is the ability to compare and communicate results obtained over a variety of hospitals or MRI devices. To fully exploit this benefit, a common presentation mode is very beneficial. For that purpose, this work encompasses a widely authored recommendation on the use of color-maps to be used for quantitative T1 and T2 maps.

Keywords

Magnetic Resonance Imaging, quantitative, relaxation parameters, sequences, sequence optimization, color map, guidelines

Citation

Fuderer, M 2025, 'Optimization of Magnetic Resonance Spin Tomography in Time-domain sequences', UMC Utrecht. https://doi.org/10.33540/3160