Alzheimer's Disease Modelling Challenge

Alzheimer's Disease Modelling Challenge

Modelling the progression of Alzheimer's disease

Summary


In this challenge, researchers propose Quantitative Templates for the Progression of Alzheimer’s disease (QT-PAD). These QT-PADs are computed from the same Alzheimer’s Disease Neuro-Imaging (ADNI) data freeze and qualitatively compared. The data freeze is named “QT-PAD Project Data”. It is available for download in the “Test Data/Data for Challenges” section of the LONI website.  This challenge is open-ended. There is no formal evaluation. Our goal is to stimulate further research and discussions.  

Background


Alzheimer’s disease (AD) results from the loss of function and death of neurons over a span of decades and is currently incurable. The availability of multiple longitudinal datasets with a profusion of simultaneous measurements allows for a novel medical research for visualizing, summarizing and aggregating these data as well as providing groupwise and personalized prediction of outcomes. Specifically, a number of research groups, see [1, 2, 3, and 4] among others, have combined heterogeneous measurements observed longitudinally and provided Quantitative templates for the progression of AD (QT-PAD). A QT-PAD is a description of the ordering and/or timing of the major changes in the path(s) to AD dementia. Contrary to conceptual or tutorial models, these templates are quantitative, derived from a cohort, and are amenable to statistical hypothesis testing. The ADNI provides the single most comprehensive public access dataset to compute QT-PADs. Up to now, QT-PADs have been computed by a number research groups using various data sets and data freeze. In order to better evaluate the pros and cons of each methodology, and to better understand the progression of AD, we offer to researchers the possibility to apply these methods to a single, common, shared data freeze. We expect that several research groups that are developing QT-PAD methodologies will use this data freeze for their work and compare their results with others.

Method


The “QT-PAD Project Data” data freeze is a subset of ADNI 1/Go/2 cohort. 
After consulting with a group of experts, we have selected the following outcomes: 
1. ADAS13
2. CDRSB
3. RAVLT.learning
4. MMSE
5. FAQ
6. FDG PET
7. Amyloid PET
8. CSF ABETA
9. CSF TAU
10. CSF PTAU
11. FS WholeBrain
12. FS Hippocampus
13. FS Entorhinal
14. FS Ventricles
15. FS MidTemp
16. FS Fusiform
And the covariates: age, APOE4 (yes/no), GENDER, EDU. 
Detailed information about these mesurements are found in the  ADNI documents and protocols
The data-freeze has been downloaded from the LONI website on June 29th 2017. It is available as a csv file. 
Let us know of any concern by sending an email to Bruno M. Jedynak at bruno.jedynak@pdx.edu

Get Involved!

Sign up to our newsletter

References
1. Donohue, Michael C., et al. "Estimating long-term multivariate progression from short-term data." Alzheimer's & Dementia 10.5 (2014): S400-S410.
2. Jedynak, Bruno M., et al. "A computational neurodegenerative disease progression score: Method and results with the Alzheimer's disease neuroimaging initiative cohort." Neuroimage 63.3 (2012): 1478-1486.
3. Schmidt-Richberg, Alexander, et al. "Multi-stage biomarker models for progression estimation in Alzheimer’s disease." International Conference on Information Processing in Medical Imaging. Springer, Cham, 2015.
4. Young, Alexandra L., et al. "A data-driven model of biomarker changes in sporadic Alzheimer's disease." Brain 137.9 (2014): 2564-2577.

Share by: