Methods of spatial normalization of positron emission tomography images

Abstract

An adaptive template image for registering a PET or a SPECT image includes a template image model including variability of values for each voxel in a template image according to one or more control parameters.

What is claimed is: 
     
         1 . An adaptive template image to be registered to a position emission tomography (PET) or a single photon emission tomography (SPECT) image during an image registration process, the adaptive template image comprising:
 a template image model wherein values associated with voxels in the adaptive template image vary according to one or more control parameters, and the adaptive template image is altered during the image registration process;   wherein variability of values for each voxel ranges from a value corresponding to the normal level of uptake of an imaging agent to an abnormal level of uptake of the imaging agent; and   
       wherein the value of the voxels are defined by:
   I i   template =I i   slope x, 
 wherein I i   template  is the value of the voxel i in the adaptive template image, I i   0  represents a value corresponding to a normal level of uptake of an imaging agent within that voxel, I i   slope  corresponds to the maximum additional amount of uptake of the imaging agent corresponding to an abnormal (high) level of uptake of the imaging agent in that voxel with respect to a control parameter x. 
 
     
     
         2 . The adaptive template image of  claim 1 , wherein the level of uptake is measured in gray matter of a brain and the imaging agent comprises an amyloid imaging agent. 
     
     
         3 . The adaptive template image of  claim 1 , wherein the variability of values for each voxel ranges from a value corresponding to a normal level of amyloid in grey matter to a raised amyloid level in grey matter. 
     
     
         4 . The adaptive template of  claim 1 , wherein the imaging agent comprises a dopamine transport agent and the level of uptake is determined for a striatum. 
     
     
         5 . The adaptive template image of  claim 1 , wherein the control parameters are selectable so as to alter the value of voxels in the adaptive image template. 
     
     
         6 . The adaptive template image of  claim 1 , wherein I 0  and I slope  are determined using regression analysis of multiple images spanning different levels of disease state. 
     
     
         7 . A method of registering a PET image to an adaptive template image during an image registration process, the method comprising:
 spatially normalizing an individual PET image to the adaptive template image located in a reference space;   comparing the spatially normalized individual PET image to the adaptive template image to determine whether both images are sufficiently converged; and   altering a parameter controlling the adaptive template image wherein values for each voxel in the adaptive template image vary, and altering parameters controlling a spatial transformation for the individual PET image, so as to render an altered template image and an updated spatially transformed individual PET image during the image registration process when the comparing step does not determine that the spatially normalized individual PET image is sufficiently converged to the adaptive template image.   
     
     
         8 . The method of  claim 7 , wherein the altering step further comprises determining whether the spatially normalized individual PET image is maximally converged to the adaptive template image. 
     
     
         9 . The method of  claim 7 , wherein said spatial normalizing step includes the steps of:
 performing linear and non-linear transformations on the individual PET image to find a match between the adaptive template image and the individual PET image, and wherein said linear transformation includes one or more of the group consisting of translation, rotation, and scaling.   
     
     
         10 . A method of constructing an adaptive template image for registering a PET image during an image registration process, the method comprising:
 calculating a mean image from a set of input PET images showing uptake of an imaging agent in a region of interest, the set of input PET images having been transformed into a reference space; and   calculating a regression model for each voxel in the set of input PET images having been transformed into the reference space, where a dependent variable represents voxel intensities in the input PET images and is regressed on some variable which, for each input image, represents a “true” value of where on a scale from a normal level of uptake to a high level of uptake of the imaging agent the voxel is located,   wherein intensities in the adaptive template image are continuously altered during the image registration process to make the adaptive template image converge to optimally fit the PET image being registered.   
     
     
         11 . A method of  claim 10 , wherein the normal level of uptake corresponds to Aβ− and the high level of uptake corresponds to Aβ+ and the imaging agent is an amyloid imaging agent. 
     
     
         12 . A positron emission tomography (PET) system comprising:
 a storage device; and   an image processor for registering a PET image to an adaptive template image during an image registration process, the image processor programmed to:
 spatially normalize an individual PET image to an adaptive template image located in a reference space; 
 compare the spatially normalized individual PET image to the adaptive template image to determine whether both are sufficiently converged; and 
 alter a parameter controlling the template image wherein values for each voxel in the adaptive template image vary, and alter parameters controlling a spatial transformation for the individual PET image so as render an altered template image and an updated spatially transformed individual PET image during the image registration process when the comparing step does not determine that the spatially normalized individual PET image is sufficiently converged to the adaptive template image. 
   
     
     
         13 . The positron emission tomography (PET) system of  claim 12  further comprising: a detector for detecting positron emissions from a brain of a subject, wherein the detector generates signals representing the positron emissions that are stored on the storage device.