DIMINISHING VARIANCE ALGORITHM RESPIRATORY COMPENSATION FOR FAST SPIN ECHO

点击翻译按钮获取中文摘要

Introduction: Fast Spin Echo (FSE) is incompatible with ordered phase encoded respiratory compensation. For patients incapable of breath-holds or when additional SNR or resolution is desired, signal averaging, respiratory triggering and multiple breath-holds techniques are available. We propose an easily implemented method that requires no patient cooperation. We have modified the diminishing variance algorithm (DVA) to work on long TR, multiple slice sequences without a navigator echo. Methods: We use pneumatic bellows to provide respiratory position information instead of navigator echos. Though navigators are more accurate, navigators leave a signal loss band across the abdomen and require additional processing time. Bellows information is available easily and at a higher sampling rate. During the first pass through the sequence, the respiratory position and its corresponding excitation are stored in a separate histogram for each slice. After completing the first pass, information from all the histograms is used to determine the subject's most common respiratory position (mode). Thereafter, the histogram of each individual slice is narrowed about the mode during a reacquisition period. Prior to each reacquisition excitation, the bellows provide a respiratory position. If this position will narrow the slice's histogram, then the slice's worst excitation is reacquired. The reacquisition period continues until all histograms are sufficiently narrow or the scan reaches a pre-defined time limit. Results: Though DVA and respiratory triggering have similar objectives, DVA guarantees the smallest variance of respiratory positions for a fixed scan time. DVA also allows a consistent TR independent of respiration rate. Conclusions: DVA compensation for FSE reduces motion artifacts while efficiently using scan time. It is easy to implement, works with uncooperative subjects, and does not limit other gating or contrast parameters.