|
|
|
|
|
We analyse large datasets of neuroimaging and behavioral data in population cohorts to relate variations across participants in behavioral performance (such as motor, memory, attention,...but also emotion, personality traits, etc...) to variations in their brain functional connectivity as measured with functional Magnetic Resonance Imaging (fMRI) signal at rest (i.e. resting-state functional connectivity).
|
|
|
|
|
|
Our project aim to investigate the influencing factors on the accuracy, reliability and biological interpretability of the prediction. Those factors include:
|
|
|
|
|
|
- pre-processing procedures to clean the MRI data,
|
|
|
|
|
|
- confounds removal (such as age, total brain volume, ...)
|
|
|
|
|
|
- feature selection and data compression approaches (such as a priori-selection of connectivity edges and nodes' size)
|
|
|
|
|
|
- machine learning algorithms for prediction (RVM, Elastic Nets, ...)
|
|
|
|
|
|
This work will support the development of methods for cognitive neuroscience questions such as understanding the relationship between brain regions and behavioral functions.
|
|
|
|
|
|
The future perspective of our work will include the prediction of symptoms in population with neuropsychiatric disorders (such Alzheimer's disease, schizophrenia,...).
|
|
|
|
|
|
Investigators:
|
|
|
|
|
|
Jianxiao Wu, PhD student, INM-7, <j.wu@fz-juelich.de>
|
|
|
|
|
|
Sarah Genon, Group Leader, INM-7, <s.genon@fz-juelich.de> |