GRAPPA 기법을 이용한 VIBE Sequence검사에서 GRAPPA factor와 ACS Line의 수 변화에 따른 화질 평가

2016 
Purpose : GRAPPA factor determines the amount of missing signal in the k-space and reduces the scan time in the GRAPPA technique. However, there are restrictions on decrease of SNR and increase of artifact as GRAPPA factor increased. Therefore, in this study, analyze and assess the correlation between the image quality and parameters by changing GRAPPA factor and ACS line number. Materials and Methods : The VIBE sequence with the number of GRAPPA factor 2, 3, 4, 5 increasing ACS line number 20, 40, 60, 80, 100, 120 was used. SNR measured to evaluation of image quality. Similarity evaluation was evaluated between images that used the GRAPPA technique and not used. In addition, the correlation analysis was carried out between ACS line and SNR, ACS line and the SSIM when examined in GRAPPA factor 2, 3, 4, 5. Results : As GRAPPA factor increased, SNR reduced and SNR was increased with the greater the number of ACS line under same GRAPPA factor. In the evaluation of the similarity, SSIM decreased by increasing GRAPPA factor and SSIM increasing the number of ACS Line under same GRAPPA factor is closer to 1. In the correlation analysis between the number of ACS line and SNR, Pearson correlation coefficient was 0.291 in GRAPPA factor 2 and 0.458 in GRAPPA factor 3, and 0.779 in GRAPPA factor 4 and 0.88 in GRAPPA factor 5. In the correlation analysis between the number of ACS line and SSIM, Pearson correlation coefficient was 0.787 in GRAPPA factor 2 and 0.792 in GRAPPA factor 3 and 0.918 in GRAPPA factor 4 and 0.946 in GRAPPA factor 5. Conclusion : Increasing GRAPPA factor in the GRAPPA technique can reduce scan time. however, reduce SNR of the images and increase artifact. Therefore, using the maximum ACS line in proper scan time in the GRAPPA technique can increase SNR and avoid artifact. in addition, improvement increase by increasing the number of ACS Line when GRAPPA factor is higher.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    0
    References
    0
    Citations
    NaN
    KQI
    []