Parsing local signal evolution directly from a single-shot MRI signal: a new approach for fMRI.

摘要: In this work a new single-shot MRI method, parameter assessment by retrieval from signal encoding (SS-PARSE), is introduced. This method abandons fundamental simplifying assumption that used in conventional methods. Established methods implicitly assume the local intrinsic does not change its amplitude or phase during acquisition, even though these changes may be substantial, especially relatively long signals image acquisitions. SS-PARSE, on other hand, acknowledges decay and evolution, models each datum as sample (k,t)-space rather than k-space. Because of more accurate model, SS-PARSE promises improved performance terms accuracy robustness, but requires intensive reconstruction computations. The theoretical properties are discussed, simulation results presented demonstrate robust measurements relaxation rate associated with brain activation functional (fMRI), freedom geometric errors due to off-resonance frequencies, better tolerance large susceptibility gradients occur naturally parts brain. addition, technique has potential assess nonexponential behavior signal.