Home > Publications database > The challenges and prospects of brain-based prediction of behaviour |
Journal Article | FZJ-2023-03484 |
; ; ; ;
2023
Nature Research
London
This record in other databases:
Please use a persistent id in citations: doi:10.1038/s41562-023-01670-1 doi:10.34734/FZJ-2023-03484
Abstract: Relating individual brain patterns to behaviour is fundamental in systemneuroscience. Recently, the predictive modelling approach has becomeincreasingly popular, largely due to the recent availability of large opendatasets and access to computational resources. This means that we can usemachine learning models and interindividual differences at the brain levelrepresented by neuroimaging features to predict interindividual differencesin behavioural measures. By doing so, we could identify biomarkers andneural correlates in a data-driven fashion. Nevertheless, this budding fieldof neuroimaging-based predictive modelling is facing issues that may limitits potential applications. Here we review these existing challenges, as wellas those that we anticipate as the field develops. We focus on the impactsof these challenges on brain-based predictions. We suggest potentialsolutions to address the resolvable challenges, while keeping in mind thatsome general and conceptual limitations may also underlie the predictivemodelling approach.
![]() |
The record appears in these collections: |