Abstract

The purpose of this study was to determine the feasibility of database prescan as an alternative to conventional autoprescan in pediatric brain MRI. Autoprescan parameters [receiver levels and transmit gain (TG)] were analyzed prospectively in 236 pediatric brain MRI studies. Paired t test and linear regression analysis were performed to determine predictability of autoprescan parameters by database-generated parameters. Signal-to-noise ratio, image quality, and potential time efficiency of database-generated parameters were assessed. No statistical difference (P = .13) and a high correlation between the TG of the axial fast spin echo (FSE) proton density (PD) and axial FSE T2-weighted sequences (r = .92) was seen. Strong correlations were noted between the TG of the sagittal T1-weighted and the TG of the axial FSE PD (r = .79), axial FSE T2-weighted (r = .81), and contrast-enhanced T1-weighted (r = .78) sequences. The receiver levels did not change significantly between sequences. Quantitative and qualitative analyses revealed no differences in the signal-to-noise ratios of the autoprescan and the database-predicted prescan parameters. Implementation of database prescan could improve time efficiency by 28 to 33%. Autoprescan parameters can be predicted by using database-generated information while preserving the diagnostic image quality of the study. Incorporation of database prescan into commercial MRI systems could improve MRI time efficiency and patient throughput.

View details for Web of Science ID A1997WP11800032

View details for PubMedID 9090606