Spatial Normalization Method Based on Discrete Cosine Transform

摘要

The signal from the measurement of brain activation using functional resonance imaging (fMRI) is too small [1] to be determined correctly, Therefore, the statistical methods usually used in the data analysis for fMRI with multiple subjects and multiple measurements for each subject. Therefore, it is necessary to make the average for these data from different subjects or one subject in different measurement. Just the anatomical differences between individuals make a certain uncertainty while people want to get the average data from the subjects. To resolve this problem, spatial normalization need to be implemented. This paper introduced some methods of spatial normalization and presented a more efficient one based on discrete cosine transform.