Related Experiment Video
Updated: Nov 4, 2025
07:05
A Protocol for the Administration of Real-Time fMRI Neurofeedback Training
Published on: August 24, 2017
11.2K
Predictors of real-time fMRI neurofeedback performance and improvement - A machine learning mega-analysis
Amelie Haugg1, Fabian M Renz2, Andrew A Nicholson2
1Department of Psychiatry, Psychotherapy and Psychosomatics, Psychiatric University Hospital, University of Zurich, Switzerland; Faculty of Psychology, University of Vienna, Austria.
Neuroimage
|May 28, 2021
Summary
Real-time fMRI neurofeedback training success is influenced by study design and participant type. Pre-training familiarization runs and patient populations showed improved neurofeedback performance, guiding future interventions.
Area of Science:
- Neuroscience
- Cognitive Science
- Medical Imaging
Background:
- Real-time fMRI neurofeedback enables self-regulation of brain activity, showing promise for behavioral and clinical improvements.
- However, variable participant success rates limit neurofeedback's clinical efficacy.
- Identifying factors influencing neurofeedback performance is crucial for optimizing interventions.
Purpose of the Study:
- To investigate design-specific, region-specific, and subject-specific factors impacting neurofeedback performance using a machine learning approach.
- To identify key predictors of successful neurofeedback training across diverse studies.
Main Methods:
- A big data machine learning analysis was performed on data from 608 participants across 28 independent real-time fMRI neurofeedback experiments.
- Investigated 20 potential influencing factors, including pre-training runs, patient vs. healthy participant status, and region of interest specifics.
Main Results:
- Machine learning identified two significant factors influencing neurofeedback performance with 60% accuracy.
- Inclusion of a pre-training no-feedback run and training patients (vs. healthy controls) were associated with better neurofeedback performance.
Conclusions:
- Pre-training familiarization and patient motivation/experimental design may enhance neurofeedback success.
- Findings offer guidance for designing more effective neurofeedback studies and interventions, particularly for clinical populations.
- Promoting open science and data sharing can accelerate data-driven recommendations.
