Based on this figure, it can be anticipated that at UHF, experiments will be limited by reduced excitation or inversion bandwidths. 16A-16C shows examples of simulated profiles for excitation and inversion pulses at three different RF field strengths. The increased pulse length may also cause magnetization losses due to the intrinsic relaxation mechanisms of the system under examination. For instance, to achieve the desired nutation angle, low-power RF pulses generally need to be applied for a longer time, which results in a smaller excitation band and leads to phase distortions and lower peak intensities. However, reducing RF pulse power affects the bandwidth as well as the spectral quality. Hence, the design of low-power RF pulses provides a solution to the bandwidth problem for UHF. As the bandwidth of RF pulses becomes smaller the power decreases and ultra-broad band pulses become equivalent to low-power pulses with regular bandwidth. A possible solution is to use RF pulses with ultra-broad operational bandwidths. Such user input 114 can include parameters or other data provided by a user to the UPF module 108, which may be provided via a user interface such as a graphical user interface. In some implementations, user input 114 is provided to the UPF module 108. The UPF module 108 can then output RF pulse data 112, which may include the RF pulse parameters associated with the user-selected RF pulse. The optimal phase surface and the CIA files are provided to the UPF module 108 to allow user selection of an RF pulse profile from the optimal phase surface. The CIA module 106 indexes the optimal phase surface and creates index files, which may be referred to as CIA files. Regarding the former implementations, as one example, the stopping criterion may be when the digitization limit of a given resolution of the pulse is reached. This iterative updating and optimization of the IFSA module 102 and ISCA module 104 creates an optimal phase surface, which in some instances is transferred to the CIA module 106 when the iterative steps reach a stopping criterion, and in some other instances may be input to a suitably trained neural network to generate output as one or more designed RF pulse shapes. In these instances, the ISCA module 104 further updates the parameter set and provides the further updated parameter set to another implementation of the IFSA module 102. The ISCA module 104 then receives the updated parameter set from the IFSA module 102 when a fidelity check is not satisfied. In general, the IFSA module 102 receives input data 110, such as an initial parameter set defining an initial RF pulse profile, that is processed to generate an updated parameter set defining one or more updated RF pulse profiles. Furthermore, arbitrary RF target patterns cannot be solved under these idealized methods. However, hardware limitations often reduce the performance of these pulse sequences, and often jeopardize the outcome of these experiments. Each of these methods relies on making assumptions to simplify the design of RF pulses into a practical matter. The SLR method simplifies the Bloch equations to the design of two polynomials, which can be solved using digital filter design algorithms. One method in particular is the Shinnar-Le Roux (SLR) pulse design. Several alternative methods have been proposed for large flip angles, including Bloch equation approximations, numerical simulators, and control theory methods. These approximations work poorly for large flip angles and spin echo pulses.
In the instance where there is a constant gradient, the RF excitation waveform can be approximated by the inverse Fourier Transform of the desired frequency profile using a linear excitation k space analysis. One of the most common RF pulse designs in MRI is the small flip-angle, where magnetization is tipped down from equilibrium to the transverse plane by a small angle of (<30°).
For example, the standard practice of designing RF pulses is to start with a desired bandwidth and select the predefined shapes from a library by modifying the amplitude and duration by assuming idealized conditions or by making several approximations.
0 kommentar(er)
