Skip to content
Open Imaging Index
CT MRI PET Brain

OASIS-3

OASIS-3: Longitudinal Multimodal Neuroimaging, Clinical, and Cognitive Dataset for Normal Aging and Alzheimer Disease

Retrospective longitudinal collection from the WUSTL Knight ADRC with MRI, amyloid, FDG and tau PET and CT for about 1,378 cognitively normal and cognitively declining adults aged 42 to 95, with clinical and cognitive data.

Go to dataset Edit Report

Overview

OASIS-3 is the third release of the Open Access Series of Imaging Studies. It gathers about three decades of research imaging and clinical follow-up from several studies run at the Knight Alzheimer Disease Research Center of Washington University in St. Louis, and is used for work on normal ageing, preclinical Alzheimer's disease and the link between amyloid, tau, brain structure and cognition.

Composition

The current release lists 1,378 participants: 755 cognitively normal adults and 622 with some degree of cognitive decline, aged 42 to 95. Many were followed for years, giving 2,842 MR sessions, 2,157 PET sessions with PiB, florbetapir (AV45) or FDG, and 1,472 CT sessions. A separately requested sub-project, OASIS-3_AV1451, adds 451 tau PET sessions. Clinical Dementia Rating, neuropsychological tests, UDS forms and other clinical data come with the images. Identifiers were replaced and all dates are expressed as days since study entry.

Acquisition

All MRI was acquired on Siemens systems: a 1.5 T Vision, two 3 T TIM Trio scanners and a 3 T Biograph mMR PET-MR. Sequences include T1w, T2w, FLAIR, ASL, SWI, time-of-flight angiography, resting-state BOLD and diffusion. SWI, meant for spotting microhaemorrhages, comes as magnitude, phase, minimum intensity projection and combined SWI images; the data dictionary lists 1,231 SWI scans, almost all at 3 T and from the 2018 release. T2star is listed as a scan type, but no count is published. PET was done on an ECAT HR+, a Biograph 40 PET/CT and the Biograph mMR. Images are converted to NIfTI with dcm2niix and organised in BIDS, with acquisition parameters in JSON sidecars.

Annotations

No manual labels. Many MR sessions include FreeSurfer segmentations and cortical measures, and PET sessions come with regional outputs of the PET Unified Pipeline (PUP); amyloid Centiloid values are also provided.

Known limitations

  • Data come from several studies with different visit schedules, so follow-up length and the set of scans vary by participant.
  • Scanners changed over the years (1.5 T to 3 T, PET-only to PET-MR), which affects longitudinal comparisons.
  • The data descriptor preprint describes an earlier, smaller release (1,098 participants); the website gives slightly different totals for the current release (1,378 in the text, 1,379 in the counter).
  • The sex and age distribution of the current release is not published on the website.

Cohort

Aggregate numbers from the sources below. Bars are relative to the 1,378 subjects.

Modality

studies

  • MRI 2,842
  • PET 2,157
  • CT 1,472

Contrast / sequence

scans, values can overlap

  • Susceptibility-weighted 1,231

PET tracer

studies

  • [18F]flortaucipir 451

Condition

subjects, values can overlap

  • Healthy control 755 55%
  • Cognitive impairment 622 45%

License and access

Our reading of the license, not legal advice. Before you use the data, read the original license and confirm that your use is allowed. We take no responsibility for how you use a dataset. Full disclaimer

Access
Application

A research proposal is reviewed and approved

Access page

OASIS Data Use Agreement

Non-commercial academic research only, after an application reviewed by OASIS staff. No sharing with anyone, collaborators included, without written permission; data must be destroyed or returned when the research ends. Publications must acknowledge OASIS, and work using AV45 or AV1451 PET must go to Avid Radiopharmaceuticals for review 30 days in advance.

The tau PET sub-project OASIS-3_AV1451 needs a separate data request.

Original license text Version read: OASIS Data Use Agreement Terms, July 2025 Checked 2026-10-07

What you can do

  • No
  • Not stated
  • Conditional
  • Conditional

What you can share

  • No
  • Not stated
  • Share trained models Not stated

What you must do

  • Yes
  • Share alike No
  • Yes
  • Ethics approval No
  • Conditional
  • Release code No
  • No
  • Yes

Limits

  • Yes
  • No

Commercial license: Conditional Commercial use needs explicit, prior written authorization from OASIS. No commercial license offer is published.

Citation

LaMontagne PJ, Benzinger TLS, Morris JC, et al. OASIS-3: Longitudinal Neuroimaging, Clinical, and Cognitive Dataset for Normal Aging and Alzheimer Disease. medRxiv (2019). doi:10.1101/2019.12.13.19014902

All numbers

Every number on this page, as stored in stats.csv, with its source.

MeasureBreakdownValueSource
Subjectstotal
summary text says 1378 participants; the counter on the same page shows 1379
1,378oasis3-page
Subjectscondition=healthy
cognitively normal adults
755oasis3-page
Subjectscondition=cognitive_impairment
"individuals at various stages of cognitive decline"
622oasis3-page
Studiesmodality=MR
MR sessions
2,842oasis3-page
Studiesmodality=PT
PET sessions (PIB, AV45 and FDG); the 451 tau PET sessions are counted separately
2,157oasis3-page
Studiesmodality=CT
CT sessions
1,472oasis3-page
Studiestracer=flortaucipir
tau PET sessions in sub-project OASIS-3_AV1451
451oasis3-page
Minimum agetotal
age range given for all participants
42oasis3-page
Maximum agetotal 95oasis3-page
Studiescontrast=swi
MR sessions with SWI in the earlier release described by the preprint (2168 MR sessions, 1098 participants): 2 at 1.5 T and 1217 at 3 T
1,219lamontagne2019
Table 3
Scanscontrast=swi
sum of the three disjoint rows of the scan type inventory: 2 at 1.5 T, 1217 at 3 T in the 2018 release and 12 at 3 T in the 2022 release
1,231oasis3-data-dictionary
Table 4 (p. 12)

Sources

The keys used in the table above.