Publication
A novel fast 3D-31P-B+1 mapping method called the frequency-selective Double-AngleMethod (fDAM) has been developed. This method integrates a 3D weighted stack of spiral gradient echo acquisitions with a frequency-selective pulse to efficiently map B+1 fields based on the phosphocreatine (PCr) signal at 7 T. The protocol has been optimized using simulations and validated through both phantom experiments and skeletal muscle applications. Results showed that fDAMnot only achieves a high correlation (r = 0.94) with the classical DAM (cDAM) but also provides faster and more extensive coverage. 3D-31P-B+1 mapping of the human calf muscle was completed in just 11min using fDAM compared to 24 minutes using cDAM. Furthermore, the first full-brain 31P 3D B+1 mapping was achieved in 13 minutes using a 1 Tx/32 Rx coil. These findings show that fDAM is able to provide B+1 mapping for accurate transmit field correction within the same scanning session of other 31P-MRS/I applications. A new magnetization transfer (MT)-31P-Magnetic resonance fingerprinting (MRF) approach was developed to measure the creatine kinase (CK) metabolic reaction rate (kCK) between PCr and adenosine triphosphate (ATP) in the human brain. This method extends the MRF framework to overcome conventional challenges in 31P-Measurement, such as long scan times and high specific absorption rates. A novel nested iteration interpolation method (NIIM) was introduced to efficiently handle the growing complexity of the multiparametric dictionary matching process. The combination ofMT-31P-MRF with NIIM allows for accurate estimations of the longitudinal relaxation times T1PCr and T1ATP, ATP/PCr concentration ratios Cr, B1, off resonance foff and kCK, showing excellent agreement with existing techniques like the exchange kinetics by band inversion transfer (EBIT) method and values reported in literature. This method demonstrated a significant reduction in scan time, from 17 minutes 4 seconds using EBIT to just 4 minutes 15 seconds with MT-31P-MRF, while maintaining similar or better reproducibility (with coefficients of variation less than 12%). Further advancing this approach, the MT-31P-MRF was extended to perform 3D CreatineKinase Imaging (CKI), delivering the first whole-brain kCK maps. Whole-brain CKI were acquired on a clinical 7T MRI scanner, generating high-resolution CK ratemaps in 49minutes 30 seconds. Further, within-session reproducibility assessment showed amean coefficient of variation below 11% and suggested that scans could be completed within 25minutes. In addition to kCK mapping, CKI provided simultaneous mapping of Cr, T1PCr values, and phosphorus-based B0 maps with a spatial resolution of 7.2 × 7.2 × 20mm3. In a proof-of-concept functional study, functional CKI (fCKI) was used to demonstrate CK activation in the brain's visual cortex in response to visual stimulation, revealing a mean 17% increase in CK reaction rates in the visual cortex. This represents the first 3D CK activation