Brain amyloid PET processing system and operation method thereof and non-transitory computer readable medium

An operation method of a brain amyloid PET processing system, and the operation method comprising steps of:

• processing an amyloid PET image to obtain a smoothed amyloid PET image of a normalized brain space;
• processing an MRI to obtain a whole brain white matter mask and a whole brain gray matter mask of the normalized brain space;
• extracting a whole brain white matter amyloid PET image from the smoothed amyloid PET image in a range of the whole brain white matter mask of the normalized brain space, and calculating an uptake value with a preset maximum ratio in the whole brain white matter amyloid PET image;
• and
• in the smoothed amyloid PET image of the normalized brain space, marking and counting one or more voxels in a range of the whole brain gray matter mask, wherein each voxel uptake value of the one or more voxels is greater than the uptake value of the preset maximum ratio of the whole brain white matter amyloid PET image, and using the one or more voxels for an interpretation training and a test of a classification of a machine learning,
• wherein the step of processing the MRI to obtain the whole brain white matter mask and the whole brain gray matter mask of the normalized brain space comprises:
• shifting an origin point of the MRI to an anterior commissure in the MRI;
• performing a brain tissue segmentation on the MRI to obtain a whole brain gray matter density image and a whole brain white matter density image, wherein the MRI belongs to an individual brain space;
• converting the MRI of the individual brain space into the normalized brain space through a nonlinear deformation, so as to generate a nonlinear deformation conversion parameter;
• converting the whole brain gray matter density image and the whole brain white matter density image to the normalized brain space through the nonlinear deformation conversion parameter to generate the whole brain gray matter density image of the normalized brain space and the whole brain white matter density image of the normalized brain space;
• and binarizing the whole brain gray matter density image of the normalized brain space and the whole brain white matter density image of the normalized brain space respectively by a threshold value to generate the whole brain gray matter mask of the normalized brain space and the whole brain gray matter mask of the normalized brain space.

A non-transitory computer readable medium to store a plurality of instructions for commanding a computer to execute an operation method, and the operation method comprising steps of:

• processing a amyloid PET image to obtain a smoothed amyloid PET image of a normalized brain space;
• processing an MRI to obtain a whole brain white matter mask and a whole brain gray matter mask of the normalized brain space;
• extracting a whole brain white matter amyloid PET image from the smoothed amyloid PET image in a range of the whole brain white matter mask of the normalized brain space, and calculating an uptake value with a preset maximum ratio in the whole brain white matter amyloid PET image;
• and
• in the smoothed amyloid PET image of the normalized brain space, marking and counting one or more voxels in a range of the whole brain gray matter mask, wherein each voxel uptake value of the one or more voxels is greater than the uptake value of the preset maximum ratio of the whole brain white matter amyloid PET image, and using the one or more voxels for an interpretation training and a test of a classification of a machine learning,
• wherein the step of processing the MRI to obtain the whole brain white matter mask and the whole brain gray matter mask of the normalized brain space comprises:
• shifting an origin point of the MRI to an anterior commissure in the MRI;
• performing a brain tissue segmentation on the MRI to obtain a whole brain gray matter density image and a whole brain white matter density image, wherein the MRI belongs to an individual brain space;
• converting the MRI of the individual brain space into the normalized brain space through a nonlinear deformation, so as to generate a nonlinear deformation conversion parameter;
• converting the whole brain gray matter density image and the whole brain white matter density image to the normalized brain space through the nonlinear deformation conversion parameter to generate the whole brain gray matter density image of the normalized brain space and the whole brain white matter density image of the normalized brain space;
• and binarizing the whole brain gray matter density image of the normalized brain space and the whole brain white matter density image of the normalized brain space respectively by a threshold value to generate the whole brain gray matter mask of the normalized brain space and the whole brain gray matter mask of the normalized brain space.