Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

On the logic of hypothesis testing in functional imaging.

Federico E Turkheimer1, John A D Aston, Vincent J Cunningham

  • 1Department of Neuropathology, Imperial College London, Charing Cross Hospital, St. Dunstan's Road, London, W6 8RP, UK. federico.turkheimer@imperial.ac.uk

European Journal of Nuclear Medicine and Molecular Imaging
|January 20, 2004
PubMed
Summary

This study examines statistical inference in functional imaging, highlighting how frequentist approaches diverge in multiple testing scenarios. A Bayesian formulation offers a clearer two-step inductive and deductive inference process for neuroscience data.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Blurring evidence with advocacy: a systematic review of policy recommendations for net zero.

npj environmental social sciences·2026
Same author

Peripheral inflammation is associated with reduced influx of TSPO PET tracers into the brain: insights from a non-invasive mapping methodology.

Brain, behavior, and immunity·2026
Same author

Network-based disease fingerprinting with neuroinflammation PET imaging.

Journal of neuroinflammation·2026
Same author

Normative modeling of choroid plexus volume across the adult lifespan: validation in multiple sclerosis and depression.

Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference·2025
Same author

Advancing Generalisable Neural Network-Based PET Quantification: A Multicenter [<sup>11</sup>C]PBR28 study.

Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference·2025
Same author

Immune alterations in schizophrenia and the effects of a therapeutic antibody: a neuroimaging study.

Brain : a journal of neurology·2025

Area of Science:

  • Neuroimaging
  • Statistical Inference
  • Computational Neuroscience

Background:

  • Functional imaging relies heavily on statistical methods for data analysis.
  • Current statistical algorithms in neuroimaging may unduly influence data interpretation.
  • Understanding the logical underpinnings of these statistical approaches is crucial for valid inference.

Purpose of the Study:

  • To investigate the logical foundations of current statistical methods in functional imaging.
  • To assess the suitability of these methods for inductive inference in neuroscience.
  • To propose an improved inferential framework using a Bayesian approach.

Main Methods:

  • Review of frequentist statistical inference, including Fisherian significance testing and Neyman-Pearson hypothesis testing.

Related Experiment Videos

  • Analysis of the dissociation between these methods in the context of multiple testing problems common in functional imaging.
  • Recasting the multiple comparison problem into a multivariate Bayesian formulation.

    • Main Results:

    • Frequentist approaches, while similar in univariate cases, differ significantly when applied to the multiple testing problem in functional imaging.
    • Issues like small volume correction can cause confusion for practitioners.
    • A multivariate Bayesian formulation clarifies the inferential process into distinct inductive and deductive steps.

    Conclusions:

    • Current statistical practices in functional imaging require critical evaluation regarding their inferential capabilities.
    • A two-step Bayesian approach, combining exploratory techniques with rigorous hypothesis testing, provides a more robust framework for neuroscience data analysis.
    • This refined methodology enhances the clarity and validity of inferences drawn from neuroimaging studies.