This title appears in the Scientific Report :
2023
Please use the identifier:
http://dx.doi.org/10.1101/2023.02.18.529076 in citations.
Please use the identifier: http://dx.doi.org/10.34734/FZJ-2024-01223 in citations.
Is resting state fMRI better than individual characteristics at predicting cognition?
Is resting state fMRI better than individual characteristics at predicting cognition?
Resting state fMRI versus confounds for behavioral phenotypic predictionIs resting state fMRI better than individualcharacteristics at predicting cognition?Amir Omidvarnia1,2*, Leonard Sasse1,2, Daouia I. Larabi1,2, FedericoRaimondo1,2, Felix Hoffstaedter1,2, Jan Kasper1,2, Juergen Dukart1,2,Marvin...
Saved in:
Personal Name(s): | Omidvarnia, Amir (Corresponding author) |
---|---|
Sasse, Leonard / Larabi, Daouia I. / Raimondo, Federico / Hoffstaedter, Felix / Kasper, Jan / Dukart, Juergen / Petersen, Marvin / Cheng, Bastian / Thomalla, Götz / Eickhoff, Simon B. / Patil, Kaustubh R. | |
Contributing Institute: |
Gehirn & Verhalten; INM-7 |
Imprint: |
2023
|
DOI: |
10.1101/2023.02.18.529076 |
DOI: |
10.34734/FZJ-2024-01223 |
Document Type: |
Preprint |
Research Program: |
Neuroscientific Data Analytics and AI Neuroimaging |
Link: |
OpenAccess |
Publikationsportal JuSER |
Please use the identifier: http://dx.doi.org/10.34734/FZJ-2024-01223 in citations.
Resting state fMRI versus confounds for behavioral phenotypic predictionIs resting state fMRI better than individualcharacteristics at predicting cognition?Amir Omidvarnia1,2*, Leonard Sasse1,2, Daouia I. Larabi1,2, FedericoRaimondo1,2, Felix Hoffstaedter1,2, Jan Kasper1,2, Juergen Dukart1,2,Marvin Petersen3, Bastian Cheng3, Götz Thomalla3, Simon B. Eickhoff1,2,Kaustubh R. Patil1,21Institute of Neuroscience and Medicine, Brain & Behavior (INM-7),Research Center Jülich, Wilhelm-Johnen-Straße, Jülich, 52428, Germany2Institute of Systems Neuroscience, Medical Faculty, Heinrich HeineUniversity Düsseldorf, Moorenstr. 5, Düsseldorf, 40225, Germany3Klinik und Poliklinik für Neurologie, Kopf- und Neurozentrum, UniversityMedical Center Hamburg-Eppendorf, Hamburg, Germany*Corresponding author(s). Email(s): a.omidvarnia@fz-juelich.deAbstractChanges in spontaneous brain activity at rest provide rich informationabout behavior and cognition. The mathematical properties of resting-statefunctional magnetic resonance imaging (rsfMRI) are a depiction of brainfunction and are frequently used to predict cognitive phenotypes.Individual characteristics such as age, gender, and total intracranial volume(TIV) play an important role in predictive modeling of rsfMRI (for example,as "confounders" in many cases). It is unclear, however, to what extentrsfMRI carries independent information from the individual characteristicsthat is able to predict cognitive phenotypes. Here, we used predictivemodeling to thoroughly examine the predictability of four cognitivephenotypes in 20,000 healthy UK Biobank subjects. We extracted commonrsfMRI features of functional brain connectivity (FC) and temporalcomplexity (TC). We assessed the ability of these features to predictoutcomes in the presence and absence of age, gender, and TIV. Additionally,we assessed the predictiveness of age, gender, and TIV only. We find TC andFC features to perform comparably with regard to predicting cognitivephenotypes. As compared to rsfMRI features, individual characteristicsprovide systematically better predictions with smaller sample sizes and, tosome extent, in larger cohorts. It is also consistent across different levels ofinherent temporal noise in rsfMRI. Our results suggest that when theobjective is to perform cognitive predictions as opposed to understandingthe relationship between brain and behavior, individual characteristics aremore applicable than rsfMRI features. |