High Field In vivo 13C Magnetic Resonance Spectroscopy of Brain by Random Radiofrequency Heteronuclear Decoupling and Data Undersampling

In vivo13C magnetic resonance spectroscopy (MRS) is a unique and effective tool for studying dynamic human brain metabolism and the cycling of neurotransmitters. One of the major technical challenges for in vivo13C-MRS is the high radio frequency (RF) power necessary for heteronuclear decoupling. In the common practice of in vivo13C-MRS, alkanyl carbons are detected in the spectra range of 10-65 ppm. The amplitude of decoupling pulses has to be significantly greater than the large one-bond 1H-13C scalar coupling (1JCH = 125-145 Hz). Two main proton decoupling methods have been developed: broadband stochastic decoupling and coherent composite or adiabatic pulse decoupling (e.g., WALTZ); the latter is widely used because of its efficiency and superb performance under inhomogeneous B1 field. Because the RF power required for proton decoupling increases quadratically with field strength, in vivo13C-MRS using coherent decoupling is often limited to lowmagnetic fields [<=4 Tesla (T)] to keep the local and averaged specific absorption rate (SAR) under the safety guidelines established by the International Electrotechnical Commission (IEC) and the US Food and Drug Administration (FDA). Alternately, carboxylic/amide carbons are coupled to protons via weak long-range 1H-13C scalar couplings, which can be decoupled using low RF power broadband stochastic decoupling. Recently, the carboxylic/amide 13C-MRS technique using low power random RF heteronuclear decoupling was safely applied to human brain studies at 7T. Here, we review the two major decoupling methods and the carboxylic/amide 13C-MRS with low power decoupling strategy. Further decreases in RF power deposition by frequency-domain windowing and time-domain random under-sampling are also discussed. Low RF power decoupling opens the possibility of performing in vivo13C experiments of human brain at very high magnetic fields (such as 11.7T), where signal-to-noise ratio as well as spatial and temporal spectral resolution are more favorable than lower fields.