Methodological improvements of the Mescher-Garwood (MEGA) J-editing sequence in conjunction with adiabatic pulses

Abstract: Magnetic resonance spectroscopy (MRS) is a technique which allows to non-invasively study the metabolism of a living organism by the acquisition of localized spectra. This work is concerned with 1H-MRS of the human brain, where the detection and discrimination of metabolites is limited by the achievable spectral separation of a clinical MR scanner and by the intrinsically low sensitivity of MRS itself. The development of MRS techniques that allow or improve the detection of more metabolites is an active research field. One clinically important method is the MEscher GArwood (MEGA) J-editing scheme, which improves or enables the detection of J-coupled metabolites like lactate or γ-aminobutyric acid (GABA). This PhD thesis explores methodological improvements of the MEGA technique. In the course of this work, several limitations of the MEGA technique were identified and mitigations strategies investigated. A fundamental improvement is the application of adiabatic pulses, which are robust against inhomogeneities of the transmit field and provide high bandwidths.

At first, the interaction between MEGA editing and spatial localization was investigated. The benefits of the sLASER localization scheme based on adiabatic pulses for MEGA editing are demonstrated for the metabolite lactate by numerical simulations, phantom and in vivo experiments. The increased signal-to-noise ratio of MEGA-sLASER compared to MEGA-PRESS, which employs conventional pulses for localization, makes it more suitable to detect subtle changes in less abundant metabolites as demonstrated by the substantially improved detection of the metabolite β-hydroxybutyrate (BHB) in a subject undergoing a low-carbohydrate (ketogenic) diet. A major limitation of the MEGA scheme lies in the inflexibility to freely choose the echo time.

In order to obtain an optimal signal-to-noise ratio a specific echo time has to be employed, which is dictated by the spin system of the target metabolite. To allow for a larger variety of experiments, a modification of the MEGA scheme is presented that decouples the echo time dependence from the spin system and thus enables efficient MEGA editing at arbitrary echo times above a threshold. The modified MEGA scheme allowed to perform T2 relaxation time measurements for lactate, which unexpectedly revealed oscillations superimposed on the exponential decay curve. Analytical calculations and experiments showed that the oscillations were caused by flip angle errors of the editing pulses either due to erroneous calibration or an inhomogeneous transmit field.

Motivated by that observation, asymmetric adiabatic pulses for J-editing were explored. It is demonstrated that the proposed asymmetric adiabatic pulse offers the advantages of immunity against calibration errors or inhomogeneities of the transmit field and against frequency drifts of the main field at the expense of increased co- editing and power deposition compared to a conventional Gaussian editing pulse. Furthermore, in contrast to symmetric pulses, asymmetric (adiabatic) pulses can be paired in different configurations within the MEGA editing scheme. It is shown that different arrangements of the asymmetric adiabatic editing pulse pair enable longer or shorter J-refocusing intervals for a given echo time, which can be exploited to increase the flexibility of the MEGA scheme.

By combining the flexible MEGA scheme with the improved robustness of asym- metric adiabatic pulses, it was possible to acquire in vivo T2 relaxation curves of lactate with an almost perfect suppression of the unwanted oscillations. The pre- sented methodology is not limited to lactate and can be applied to a wide range of J-coupled metabolites. An important application for future work would be the T2 determination for other J-coupled metabolites, which is essential for absolute quantification

Angaben zum Objekt

• Standort: Deutsche Nationalbibliothek Frankfurt am Main
• Umfang: Online-Ressource
• Sprache: Englisch
• Anmerkungen: Universität Freiburg, Dissertation, 2021

Klassifikation und Themen

• Schlagwort: Editing

Beteiligte, Orts- und Zeitangaben

• Ereignis: Veröffentlichung
• (wo): Freiburg
• (wer): Universität
• (wann): 2021
• Urheber: Dacko, Michael
• Beteiligte Personen und Organisationen: Hennig, Jürgen ; Lange, Thomas ; Albert-Ludwigs-Universität Freiburg. Fakultät für Mathematik und Physik

Weitere Informationen

• DOI: 10.6094/UNIFR/222610
• URN: urn:nbn:de:bsz:25-freidok-2226103
• Rechteinformation: Kein Open Access; Der Zugriff auf das Objekt ist unbeschränkt möglich.
• Letzte Aktualisierung: 25.03.2025, 13:43 MEZ