|
|
||||||||
|
|
|||||||||
|
|
Technologist Student AbstractsTechnologist Student Posters |
1 Nuclear Department, Shang Hai Renji Hospital, Shang Hai, Shang Hai, China
2304
Objectives: To discuss the effection of the threshold correction for the results of the Statistic parameter mapping soft(SPM).
Methods: making defect models by Hoffman standard brain modal ,to compare their PET images with the PET images of the standard modal by SPM.And select the images of the patient with cerebral infarction and health persons to conducte SPM statistical comparison.
Results: The results of correction and uncorrection are different.Famiywise error correction has the more power and can get least and the smallest regions whose the value of P are higher than the threshold.The false discovery rate has unsteadily results,sometimes it is worse than the uncorrecion.
Conclusions: In SPM,FWE has the least false region and the much more reliable result than the uncorrection.
Research Support: Ge YC, Sealfon SC, Tseng CH.A Holm-type procedure controlling the false discovery rate, Statistics & Probability Letters,2007,77( 18): 1756-1762 Livio Finos , Luigi Salmaso.FDR- and FWE-controlling methods using data-driven weights, Journal of Statistical Planning and Inference, 2007, 137(12): 3859-3870
| ||||||||||||||||||||||||||||||||||||||