MRI-and-PET-data-from-ADNI

Pre-processed MRI and PET images from ADNI-1/ADNI-2(https://adni.loni.usc.edu/)

GitHib is so hard to upload large file, we upload scripts instead. Processing with FSL Software is fast enough.

Advanced Search (beta)

Description of Downloaded MRI

if exist("MPR; GradWarp; B1 Correction; N3"):
  select
elif exist("MPR-R; GradWarp; B1 Correction; N3"):
  select
elif exist("MPR; GradWarp; N3"):
  select
elif exist("MPR-R; GradWarp; N3"):
  select
 elif exist("MPR; ; N3"):
  select

For the subjects with format of "MPR; ; N3" and "MPR-R; GradWarp; N3", we don not impute additional preprocessing steps such as "GradWarp" or "B1 Correction"

Preprocessing of Downloaded MRI

Above preprocessing steps have already be done, we focus on

BET algorithm (in FSL 6.0.3) is used for brain extraction.
FLIRT linear registration algorithm (in FSL 6.0.3) is used to aligne images to Montreal Neurological Institute T1 standard template space (MNI152_T1_1mm).
Remove the voxels of zero values in the periphery of brain.
Downsample the images to the size of 76×94×76.

Description of Downloaded PET

if exist("Coreg, Avg, Std Img and Vox Siz, Uniform Resolution"):
  select
elif exist("AV45 Coreg, Avg, Std Img and Vox Siz, Uniform Resolution"):
  select

Preprocessing steps of PET images are relatively uniform, that is ideal for the inputs of models.

Preprocessing of Downloaded PET

Above preprocessing steps have already be done, we focus on

BET algorithm (in FSL 6.0.3) is used for brain extraction.
FLIRT linear registration algorithm (in FSL 6.0.3) is used to aligne images to Montreal Neurological Institute T1 standard template space (MNI152_T1_1mm).
Remove the voxels of zero values in the periphery of brain.
Downsample the images to the size of 76×94×76.

About citation

[1] C. R. Jack et al., “The Alzheimer’s disease neuroimaging initiative (ADNI): MRI methods,” J. Magn. Reson. Imag., vol. 27, no. 4, pp. 685–691, 2008.

[2] X. Gao, et al., "Task-Induced Pyramid and Attention GAN for Multimodal Brain Image Imputation and Classification in Alzheimer's Disease," IEEE Journal of Biomedical and Health Informatics, vol. 26, no. 1, pp. 36-43, Jan. 2022.