Motion Correction

In functional brain imaging a series of brain images is acquired. The time elapsed between each acquisition is usually a few seconds or less. Due to the small acquisition times required, these images usually have poor resolution. Furthermore, as the imaging parameters are tuned to highlight physiological changes (e.g. blood oxygenation) the images often have poor anatomical contrast. Extracting functional information from such a series of images is done by applying statistical time-series analysis, which assumes that the location of a given voxel within the brain does not change over time. However, there is usually some degree of subject motion within the scanner, especially when the scanning takes a long time or when clinical patients are involved. Therefore, in order to render the data fit for statistical analysis this motion must be estimated and corrected for. This is the task of motion correction methods and it is essentially a multiple-image registration task. Normally motion correction methods deal with the registration task by selecting a reference image from within the series and registering each image in turn to this fixed reference. As all images are of the same subject, using the same imaging parameters, it can be classified an intra-subject, intra-modal registration problem. Therefore, a rigid-body transformation space and intra-modal cost function can be used. Furthermore, as the values in the corrected images are important for later statistical analysis, the choice of interpolation method for the transformation of the images is of particular importance (Hajnal et al.,1995a, b).