Method for processing brain images

A method for processing brain images, wherein the method comprises the steps of:

• a) Performing a co-registration, preferably a linear co-registration with 6 degrees of freedom for a PET (Positron Emission Tomography) image of a patient, using as reference a first volumetric MR (Magnetic Resonance) image of the patient, acquired in 3D mode, generating a new PET image, i.e. having identical spatial properties to the volumetric MR image, but containing the information on the patient's brain metabolism obtained in the PET image;
• b) Calculating a first spatial transformation matrix M1, necessary for deforming the volumetric MR image acquired in 3D mode, bringing it to the shape of an MR template image, which is an image acquired in 3D mode such as an MR symmetric template, thus generating a new volumetric MR image which has the particularity that the interhemispheric groove coincides entirely with the coordinate X=0;
• c) Applying the matrix M1, obtained in the preceding step, to the PET image, generating a new PET image, having identical spatial properties to the volumetric MR image, but containing the information on the patient's brain metabolism obtained in the PET image;
• d) Determining the healthy hemisphere from the volumetric MR image using the information obtained from the electroencephalographic pattern or another technique to identify the laterality of the patient's lesion, and once the healthy hemisphere has been identified, generating an image from the dimensions of the volumetric MR image, wherein the values on the X-axis corresponding to the healthy side are 1, and the values of the injured side are 0, and this image is referred to as region of interest ROI;
• e) Multiplying the volumetric MR image element-wise with the image of the region of interest ROI, obtaining a new volumetric MR image, which corresponds only to the healthy hemisphere;
• copying the volumetric MR image and inverting it on the X-axis, generating a mirror image of the volumetric MR image, referred to as volumetric MR image;
• and adding the volumetric MR image element-wise with the volumetric MR image, generating a new volumetric MR image;
• f) Multiplying the PET image element-wise with the image of the region of interest ROI, generating a new PET image;
• copying the PET image and inverting it on the X-axis, generating a mirror image of the PET image, referred to as PET image;
• and adding the PET image element-wise with the PET image, generating a new PET image;
• g) Calculating a second transformation matrix M2, necessary for deforming the volumetric MR image to the image (M 5 R 20);
• the image obtained is referred to as MR;
• h) Applying the transformation matrix M2 to the PET image, obtaining the PET image, which contains all the metabolic information of the patient's healthy hemisphere in the PET image, adding element-wise to the PET image in the space of the volumetric MR image;
• i) Calculating a third transformation matrix M3, necessary for spatially deforming the volumetric MR image to the volumetric MR image;
• and
• j) Applying the transformation matrix M3 to the PET image, obtaining the PET image, and applying the transformation matrix M2 to the PET image, obtaining a new PET image.