How to improve parameter estimates in GLM-based fMRI data analysis: cross-validated Bayesian model averaging.

Joram Soch, Achim Pascal Meyer, John-Dylan Haynes, Carsten Allefeld

September 2017 Neuroimage

Synopsis of Social media discussions

Discussions highlight the significance of methods like cvBMA in enhancing fMRI analysis, with examples such as mentions of the 2017 NeuroImage paper and its implications for better modeling. The tone and language convey a positive outlook on the research's potential to improve scientific practice, underlining its relevance and influence in the field.

A
Agreement
 Moderate agreement

The discussions reflect a consensus that the paper offers valuable advancements in fMRI data analysis, supporting its methodologies and findings.

I
Interest
 Moderate level of interest

The posts express moderate interest, mainly through technical references and recognition of the research's relevance.

E
Engagement
 Moderate level of engagement

The posts include specific mentions of the techniques like cross-validated Bayesian model averaging and their applications, indicating active engagement.

I
Impact
 Moderate level of impact

The emphasis on the paper’s contribution to improving parameter estimation suggests perceived meaningful impact within neuroimaging research.

Social Mentions

YouTube

1 Videos

Twitter

2 Posts

Metrics

Video Views

53

Total Likes

1

Extended Reach

1,958

Social Features

3

Timeline: Posts about article

Top Social Media Posts

Posts referencing the article

Enhancing fMRI Parameter Estimates with Cross-Validated Bayesian Model Averaging

Enhancing fMRI Parameter Estimates with Cross-Validated Bayesian Model Averaging

This video discusses methods to improve fMRI data analysis using General Linear Models, focusing on cross-validated Bayesian model selection and model averaging for more accurate parameter estimation and better detection of brain activity.

Joram Soch

November 4, 2022

53 views

  • Joram Soch
    @JoramSoch (Twitter)

    In this #PaperVideo, I'm summarizing our 2017 NeuroImage paper about Bayesian model averaging for general linear models in fMRI data analysis. #cvBMA #fMRI https://t.co/ULE4rRU9yz
    view full post

    October 26, 2023

  • City Research Online
    @City_Research (Twitter)

    #openaccess New research in CRO: October 22, 2019 at 11:15AM How to improve parameter estimates in GLM-based fMRI data analysis: cross-validated Bayesian model averaging https://t.co/X7kxyioYiA
    view full post

    October 22, 2019

Abstract Synopsis

  • The paper discusses methods to improve the analysis of fMRI data using General Linear Models (GLMs), specifically through crossvalidated Bayesian model selection (cvBMS) and model averaging (cvBMA), to select or combine models for better accuracy.
  • cvBMS helps identify the best model for each group, reducing the risk of missing true effects, while cvBMA combines information from all models weighted by their probabilities to improve the estimation of key brain activity parameters.
  • The authors highlight that cvBMA is especially useful when regressors of interest are consistent across models, as it can enhance sensitivity to experimental effects compared to simply choosing the best model per subject or group.]