001034775 001__ 1034775
001034775 005__ 20250203103400.0
001034775 0247_ $$2doi$$a10.1101/2024.09.20.24314055
001034775 037__ $$aFZJ-2024-07529
001034775 1001_ $$0P:(DE-Juel1)187351$$aKomeyer, Vera$$b0$$eCorresponding author
001034775 245__ $$aCorrect deconfounding enables causal machine learning for precision medicine and beyond
001034775 260__ $$c2024
001034775 3367_ $$0PUB:(DE-HGF)25$$2PUB:(DE-HGF)$$aPreprint$$bpreprint$$mpreprint$$s1736258044_14437
001034775 3367_ $$2ORCID$$aWORKING_PAPER
001034775 3367_ $$028$$2EndNote$$aElectronic Article
001034775 3367_ $$2DRIVER$$apreprint
001034775 3367_ $$2BibTeX$$aARTICLE
001034775 3367_ $$2DataCite$$aOutput Types/Working Paper
001034775 520__ $$aArtificial intelligence holds promise for individualized medicine. Yet, predictive models in the neurobiomedical domain suffer from a lack of generalizability and replicability so that transitioning models from prototyping to clinical applications still poses challenges. Key contributors to these challenges are confounding effects; in particular the oftentimes purely statistical perspective on confounding. However, complementing these statistical considerations with causal reasoning from domain knowledge can make predictive models a tool for causal biomedical inference beyond associative insights. Such causal insights give answers to biomedical questions of how and why, arguably what most biomedical investigations ultimately seek for. Here, we suggest a 5-step approach for targeted, context-informed deconfounding. We exemplify the 5-step approach with a real-world neurobiomedical predictive task using data from the UK Biobank. The core of this approach constitutes a bottom-up causal analysis to identify a correct set of deconfounders and the appropriate deconfounding method for a given causal predictive endeavour. Using the 5-step approach to combine causal with statistical confounder considerations can make predictive models based on observational (big) data a technique comparable to Randomized Control Trials (RCTs). Through causally motivated deconfounding we aim at facilitating the development of reliable and trustworthy AI as a medical tool. In addition, we aim to foster the relevance of low performing or even null result models if they originate from a “skilful interrogation of nature”, i.e. a deconfounding strategy derived from an adequate causal and statistical analysis. Ultimately, causal predictive modelling through appropriate deconfounding can contribute to mutual recursive feedback loops of causal insights across disciplines, scales and species that enable the field to disentangle the cause-effect structure of neurobiomedical mechanisms.
001034775 536__ $$0G:(DE-HGF)POF4-5254$$a5254 - Neuroscientific Data Analytics and AI (POF4-525)$$cPOF4-525$$fPOF IV$$x0
001034775 536__ $$0G:(DE-HGF)POF4-5251$$a5251 - Multilevel Brain Organization and Variability (POF4-525)$$cPOF4-525$$fPOF IV$$x1
001034775 536__ $$0G:(GEPRIS)431549029$$aDFG project G:(GEPRIS)431549029 - SFB 1451: Schlüsselmechanismen normaler und krankheitsbedingt gestörter motorischer Kontrolle (431549029)$$c431549029$$x2
001034775 588__ $$aDataset connected to CrossRef
001034775 7001_ $$0P:(DE-Juel1)131678$$aEickhoff, Simon$$b1
001034775 7001_ $$0P:(DE-Juel1)176538$$aRathkopf, Charles$$b2
001034775 7001_ $$0P:(DE-Juel1)161406$$aGrefkes, Christian$$b3
001034775 7001_ $$0P:(DE-Juel1)172843$$aPatil, Kaustubh$$b4
001034775 7001_ $$0P:(DE-Juel1)185083$$aRaimondo, Federico$$b5$$eCorresponding author
001034775 773__ $$a10.1101/2024.09.20.24314055
001034775 8564_ $$uhttps://www.medrxiv.org/content/medrxiv/early/2024/09/23/2024.09.20.24314055.full.pdf
001034775 909CO $$ooai:juser.fz-juelich.de:1034775$$pVDB
001034775 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)187351$$aForschungszentrum Jülich$$b0$$kFZJ
001034775 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)131678$$aForschungszentrum Jülich$$b1$$kFZJ
001034775 9101_ $$0I:(DE-HGF)0$$6P:(DE-Juel1)131678$$aHHU Düsseldorf$$b1
001034775 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)176538$$aForschungszentrum Jülich$$b2$$kFZJ
001034775 9101_ $$0I:(DE-HGF)0$$6P:(DE-Juel1)161406$$aINM-3$$b3
001034775 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)172843$$aForschungszentrum Jülich$$b4$$kFZJ
001034775 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)185083$$aForschungszentrum Jülich$$b5$$kFZJ
001034775 9131_ $$0G:(DE-HGF)POF4-525$$1G:(DE-HGF)POF4-520$$2G:(DE-HGF)POF4-500$$3G:(DE-HGF)POF4$$4G:(DE-HGF)POF$$9G:(DE-HGF)POF4-5254$$aDE-HGF$$bKey Technologies$$lNatural, Artificial and Cognitive Information Processing$$vDecoding Brain Organization and Dysfunction$$x0
001034775 9131_ $$0G:(DE-HGF)POF4-525$$1G:(DE-HGF)POF4-520$$2G:(DE-HGF)POF4-500$$3G:(DE-HGF)POF4$$4G:(DE-HGF)POF$$9G:(DE-HGF)POF4-5251$$aDE-HGF$$bKey Technologies$$lNatural, Artificial and Cognitive Information Processing$$vDecoding Brain Organization and Dysfunction$$x1
001034775 9141_ $$y2024
001034775 920__ $$lyes
001034775 9201_ $$0I:(DE-Juel1)INM-7-20090406$$kINM-7$$lGehirn & Verhalten$$x0
001034775 9201_ $$0I:(DE-Juel1)INM-3-20090406$$kINM-3$$lKognitive Neurowissenschaften$$x1
001034775 980__ $$apreprint
001034775 980__ $$aVDB
001034775 980__ $$aI:(DE-Juel1)INM-7-20090406
001034775 980__ $$aI:(DE-Juel1)INM-3-20090406
001034775 980__ $$aUNRESTRICTED