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

Optimizing the fMRI data-processing pipeline using prediction and reproducibility performance metrics: I. A

Stephen Strother1, Stephen La Conte, Lars Kai Hansen

  • 1Radiology Department, University of Minnesota, USA. steve@neurovia.umn.edu

Neuroimage
|October 27, 2004
PubMed
Summary

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

The epidemiological, clinical, and laboratory features of sporadic Creutzfeldt-Jakob disease patients in China: surveillance data from 2006 to 2010.

PloS one·2011
Same author

Novel LOVD databases for hereditary breast cancer and colorectal cancer genes in the Chinese population.

Human mutation·2011
Same author

Modulating the charge-transfer enhancement in GERS using an electrical field under vacuum and an n/p-doping atmosphere.

Small (Weinheim an der Bergstrasse, Germany)·2011
Same author

Myocardial injury after ischemia-reperfusion in mice deficient in Akt2 is associated with increased cardiac macrophage density.

American journal of physiology. Heart and circulatory physiology·2011
Same author

PI3K/Akt signaling requires spatial compartmentalization in plasma membrane microdomains.

Proceedings of the National Academy of Sciences of the United States of America·2011
Same author

Using FRET-based reporters to visualize subcellular dynamics of protein kinase A activity.

Methods in molecular biology (Clifton, N.J.)·2011

Optimizing data processing pipelines is crucial for accurate human brain function insights. Both prediction and reproducibility metrics are essential for tuning pipeline components, revealing interactions that isolated testing misses.

Area of Science:

  • Neuroimaging
  • Data Science
  • Brain Function Analysis

Background:

  • Published results suggest suboptimal data processing pipelines can obscure insights into human brain function.
  • Performance metrics like prediction and Receiver Operating Characteristic (ROC) curves are vital for optimizing these pipelines.

Purpose of the Study:

  • To evaluate the importance of various data processing pipeline components for analyzing BOLD-fMRI data.
  • To demonstrate the utility of the NPAIRS split-half resampling framework for estimating prediction and reproducibility metrics.
  • To investigate the interactions between different pipeline components.

Main Methods:

  • Utilized the NPAIRS framework to estimate prediction and reproducibility metrics.
  • Tested components including interpolation, spatial smoothing, temporal detrending, and alignment on BOLD-fMRI data from 16 subjects.

Related Experiment Videos

  • Employed canonical variates analysis (CVA) for detecting large-scale brain networks, tuning the model to the data.

    • Main Results:

    • Tuning the canonical variates analysis (CVA) model and spatial smoothing were identified as the most critical processing parameters.
    • Temporal detrending proved essential for removing low-frequency trends, with optimization based on prediction metrics.
    • Higher-order polynomial alignment had minimal impact compared to affine alignment.
    • Both prediction and reproducibility metrics were necessary for optimization and yielded distinct results.

    Conclusions:

    • Optimizing data processing pipelines requires considering both prediction and reproducibility metrics, as they provide complementary information.
    • Pipeline component interactions necessitate a holistic approach to optimization, rather than isolated testing.
    • Proper pipeline optimization is critical for uncovering reliable insights into human brain function from neuroimaging data.