J.-B. Poline and B.M. Mazoyer. Enhanced detection in brain activation maps using a multi filtering approach. J. Cereb. Blood Flow Metab., 14:639--641, 1994.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....the resulting image U(s) is a 2 random field with 2 degrees of freedom. P values for local maxima of such a field can be found from Worsley et al. 1996a) 2 Scale space Our main interest here is to apply scale space searches to look for signals of any extent. This method, first introduced by Poline Mazoyer (1994), smoothes the data with a Gaussian filter of varying width or scale, thus adding one extra scale dimension to the data. The scale space data is then searched for local maxima in location as well as scale. The motivation for this comes from the Matched Filter Theorem (see for example Rosenfeld ....

Poline J.B. & Mazoyer B.M. (1994). Enhanced detection in brain activation maps using a multifiltering approach. Journal of Cerebral Blood Flow and Metabolism, 14:639-42.

....signal to noise ratio is very small and so the signal is enhanced by two methods. The first involves spatial smoothing of the images, with the amount of smoothing chosen to match the signal to be detected. Trying all possible smoothing filter widths adds an extra scale space dimension to the data (Poline Mazoyer 1994; Worsley et al. 1996b) The second involves simply repeating the experiment either on different subjects (which requires careful image registration) or on different scans within the same subject. The result is a set of dependent variables, Y, one for each voxel of the images. These are related ....

Poline J.B., and Mazoyer B.M. 1994. Enhanced detection in brain activation maps using a multifiltering approach. Journal of Cerebral Blood Flow and Metabolism, 14:639-42.

....PET images of CBF in an activation study are usually smoothed to a resolution much less than that attainable by the PET camera. In many studies the choice of this smoothing is arbitrarily fixed at a 20mm FWHM, and the resulting statistical field or parametric map is searched for local maxima. Poline and Mazoyer (1994) have proposed a 4 D search over smoothing kernel widths as well as location to find local maxima in 1 D scale space as well as 3 D location space. If the peaks are well separated this allows us to estimate the size of regions of activation as well as their location. We avoid repeating the ....

Poline J.B. and Mazoyer B.M. 1994. Enhanced detection in brain activation maps using a multifiltering approach. Journal of Cerebral Blood Flow and Metabolism 14:639-42.

....could be modified to take into account individual anatomies, but as we explained above this is not our point of view. 2. 3 Multi filtering and multi scale The idea to look at different levels of scale to search for activations has emerged and is now admitted in the field of functional imaging [Poline and Mazoyer, 1994b, Poline and Mazoyer, 1994a, Worsley et al. 1996a] The idea was first introduced by Poline for his individual method of detection, essentially to solve the problem of the choice of the filter to apply. With multi filtering no a priori on the width of this filter is needed. Moreover it has been ....

Poline, J. and Mazoyer, B. (1994b). Enhanced detection in brain activation maps using a multifiltering approach. J. of Cerebral Blood Flow and Metabolism, 14:639--642.

....binary structure mask images are smoothed to improve the local signal to noise ratio and to smooth out registration errors and anatomical variability. It is not hard to show that the optimal kernel should match the shape of the signal to be detected. Since we don t know the width of the signal, Poline Mazoyer (1994) proposed searching over kernel filter width as well as location in S, but only empirical results are given. Siegmund Worsley (1995) modeled the process as a 4D non stationary Gaussian random field and approached PfZ max zg from two different directions. The first found the expected EC of ....

Poline, J-B. & Mazoyer, B.M. (1994). Enhanced detection in brain activation maps using a multifiltering approach. Journal of Cerebral Blood Flow and Metabolism, 14:639-642.

....NeuroImage 8, 302 306. arguing that correcting for the global signal (either by ANCOVA or scaling) is undesirable when the global signal itself is correlated with the task covariates of interest. Aguirre, Zarahn, and D Esposito 1998) highpass lter: spm list 1999 match medium 3.1. 7 Filtering (Poline and Mazoyer 1994) 3.1.8 Normalization Name Description Reference Proportionaly normalized ANCOVA normalization FANOVA ( Table 3.4: Normalization methods In connection with PET studies the amount of tracer in the brain may vary. The eoeect of this has to be considered. Andersson 1997) Jesper Anderson: 12 Mar ....

Poline, J.-B. and B. M. Mazoyer (1994). Enhanced detection in brain activation maps using a multiĝltering approach. Journal of Cerebral Blood Flow and Metabolism 14, 639642.

....the dependence of the data, the less severe is the test correction. Clearly, activated region with size and shape similar to the one of the filter are best detected. Since activated regions can in principle have any size or shape, multifiltering or multi scale approaches have been investigated [11, 14]. However, the greater the filter size the less precise are the boundaries of the region. In this work, we propose an alternative approach that consists in parcelling the analyzed brain into a user defined number of regions (or parcels) The parcelling method is able to take into account ....

J.-B. Poline and B.M. Mazoyer. Enhanced detection in brain activation maps using a multi filtering approach. J. Cereb. Blood Flow Metab., 14:639--641, 1994.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC