All participants including the affected brothers BI (age 28) and BII (age 25) underwent MRI using the Human Connectome Protocol and PET for tau pathology using flortaucipir. Brother BII also underwent amyloid PET imaging using florbetapir at age 27.
For tau PET, 10.3 mCi 18F-AV-1451 was administrated through an intravenous catheter and images were obtained beginning 75 min after injection. Six frames, 5 min apart, were collected and averaged. A low-dose CT transmission scan was obtained for attenuation correction. For amyloid PET, ∼50 min after intravenous administration of 10.4 mCi of 18F-florbetapir, PET images were obtained of the brain from vertex to skull base. Low-dose CT scan was obtained over the same anatomic range for attenuation correction.
To quantify flortaucipir PET imaging, we first ran recon-all script from FreeSurfer 6.0 to get high-resolution segmentations from the T1 image. Desikan–Killiany atlas16 (link) was used to define 36 anatomical regions. PET images were then co-registered to T1 native space. The Muller-Gartner method was used to correct the partial volume effect of flortaucipir PET image using PETSurfer in FreeSurfer 6.0. Standard uptake value ratios (SUVRs) were then calculated using cerebellar grey matter from the T1 image as the reference region. Partial volume–corrected SUVR was also mapped to the cortical surface that is parcellated to 36 regions for each cerebral hemisphere.
In order to investigate the integrity of white matter associated with the F388S PSEN1 substitution, diffusion MRI was performed using the Human Connectome Protocol. T1-weighted MR image and diffusion MR image of each subject were preprocessed by Human Connectome Protocol pipeline17 (link) with version 3.27. Diffusion MRI data were acquired with two opposite phase encoding directions that are anterior to posterior (AP) and posterior to anterior (PA). Raw diffusion MRI data were corrected by topup and eddy functions in FSL to reduce the distortion caused by susceptibility-induced distortion and eddy current-induced distortion. After distortion correction, diffusion MRI data were resampled to T1 image space. The diffusion tensor model and associated eigenvalues (λ1, λ2, λ3) were estimated with the MRtrix3 software.18 (link) Fractional anisotropy, mean diffusivity (MD), radial diffusivity and axial diffusivity were then computed. These diffusivity measures were mapped back to T1 space using the linear transformation obtained from the registration between B0 image and T1 image processed by recon-all. White matter is also parcellated into 36 regions for each cerebral hemisphere using the method as cortical region parcellation. Mean values of these diffusivity measures in each white matter region are then calculated for statistical analysis.
For tau PET, 10.3 mCi 18F-AV-1451 was administrated through an intravenous catheter and images were obtained beginning 75 min after injection. Six frames, 5 min apart, were collected and averaged. A low-dose CT transmission scan was obtained for attenuation correction. For amyloid PET, ∼50 min after intravenous administration of 10.4 mCi of 18F-florbetapir, PET images were obtained of the brain from vertex to skull base. Low-dose CT scan was obtained over the same anatomic range for attenuation correction.
To quantify flortaucipir PET imaging, we first ran recon-all script from FreeSurfer 6.0 to get high-resolution segmentations from the T1 image. Desikan–Killiany atlas16 (link) was used to define 36 anatomical regions. PET images were then co-registered to T1 native space. The Muller-Gartner method was used to correct the partial volume effect of flortaucipir PET image using PETSurfer in FreeSurfer 6.0. Standard uptake value ratios (SUVRs) were then calculated using cerebellar grey matter from the T1 image as the reference region. Partial volume–corrected SUVR was also mapped to the cortical surface that is parcellated to 36 regions for each cerebral hemisphere.
In order to investigate the integrity of white matter associated with the F388S PSEN1 substitution, diffusion MRI was performed using the Human Connectome Protocol. T1-weighted MR image and diffusion MR image of each subject were preprocessed by Human Connectome Protocol pipeline17 (link) with version 3.27. Diffusion MRI data were acquired with two opposite phase encoding directions that are anterior to posterior (AP) and posterior to anterior (PA). Raw diffusion MRI data were corrected by topup and eddy functions in FSL to reduce the distortion caused by susceptibility-induced distortion and eddy current-induced distortion. After distortion correction, diffusion MRI data were resampled to T1 image space. The diffusion tensor model and associated eigenvalues (λ1, λ2, λ3) were estimated with the MRtrix3 software.18 (link) Fractional anisotropy, mean diffusivity (MD), radial diffusivity and axial diffusivity were then computed. These diffusivity measures were mapped back to T1 space using the linear transformation obtained from the registration between B0 image and T1 image processed by recon-all. White matter is also parcellated into 36 regions for each cerebral hemisphere using the method as cortical region parcellation. Mean values of these diffusivity measures in each white matter region are then calculated for statistical analysis.
Free full text: Click here
