TY - JOUR
T1 - Predictors of real-time fMRI neurofeedback performance and improvement – A machine learning mega-analysis
AU - Haugg, Amelie
AU - Renz, Fabian M.
AU - Nicholson, Andrew A.
AU - Lor, Cindy
AU - Götzendorfer, Sebastian J.
AU - Sladky, Ronald
AU - Skouras, Stavros
AU - McDonald, Amalia
AU - Craddock, Cameron
AU - Hellrung, Lydia
AU - Kirschner, Matthias
AU - Herdener, Marcus
AU - Koush, Yury
AU - Papoutsi, Marina
AU - Keynan, Jackob
AU - Hendler, Talma
AU - Cohen Kadosh, Kathrin
AU - Zich, Catharina
AU - Kohl, Simon H.
AU - Hallschmid, Manfred
AU - MacInnes, Jeff
AU - Adcock, R. Alison
AU - Dickerson, Kathryn C.
AU - Chen, Nan Kuei
AU - Young, Kymberly
AU - Bodurka, Jerzy
AU - Marxen, Michael
AU - Yao, Shuxia
AU - Becker, Benjamin
AU - Auer, Tibor
AU - Schweizer, Renate
AU - Pamplona, Gustavo
AU - Lanius, Ruth A.
AU - Emmert, Kirsten
AU - Haller, Sven
AU - Van De Ville, Dimitri
AU - Kim, Dong Youl
AU - Lee, Jong Hwan
AU - Marins, Theo
AU - Megumi, Fukuda
AU - Sorger, Bettina
AU - Kamp, Tabea
AU - Liew, Sook Lei
AU - Veit, Ralf
AU - Spetter, Maartje
AU - Weiskopf, Nikolaus
AU - Scharnowski, Frank
AU - Steyrl, David
N1 - Funding Information:
A.H. was supported by the Forschungskredit of the University of Zurich (FK‐18‐030), F.S. was supported by the Foundation for Research in Science and the Humanities at the University of Zurich (STWF‐17‐012) and the Swiss National Science Foundation (32003B_166,566, BSSG10_155,915, 100,014_178,841). KCH and CZ received funding from the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement no. 602186 and registered as preclinical trial #NCT02463136. LH was supported by the European Union’s Horizon 2020 research and innovation program under the Grant Agreement No 794395. MM was supported by the Deutsche Forschungsgemeinschaft (DFG grant Nos. 178833530 [SFB 940] and 402170461 [TRR 265]).
Publisher Copyright:
© 2021 The Author(s)
PY - 2021/8/15
Y1 - 2021/8/15
N2 - Real-time fMRI neurofeedback is an increasingly popular neuroimaging technique that allows an individual to gain control over his/her own brain signals, which can lead to improvements in behavior in healthy participants as well as to improvements of clinical symptoms in patient populations. However, a considerably large ratio of participants undergoing neurofeedback training do not learn to control their own brain signals and, consequently, do not benefit from neurofeedback interventions, which limits clinical efficacy of neurofeedback interventions. As neurofeedback success varies between studies and participants, it is important to identify factors that might influence neurofeedback success. Here, for the first time, we employed a big data machine learning approach to investigate the influence of 20 different design-specific (e.g. activity vs. connectivity feedback), region of interest-specific (e.g. cortical vs. subcortical) and subject-specific factors (e.g. age) on neurofeedback performance and improvement in 608 participants from 28 independent experiments. With a classification accuracy of 60% (considerably different from chance level), we identified two factors that significantly influenced neurofeedback performance: Both the inclusion of a pre-training no-feedback run before neurofeedback training and neurofeedback training of patients as compared to healthy participants were associated with better neurofeedback performance. The positive effect of pre-training no-feedback runs on neurofeedback performance might be due to the familiarization of participants with the neurofeedback setup and the mental imagery task before neurofeedback training runs. Better performance of patients as compared to healthy participants might be driven by higher motivation of patients, higher ranges for the regulation of dysfunctional brain signals, or a more extensive piloting of clinical experimental paradigms. Due to the large heterogeneity of our dataset, these findings likely generalize across neurofeedback studies, thus providing guidance for designing more efficient neurofeedback studies specifically for improving clinical neurofeedback-based interventions. To facilitate the development of data-driven recommendations for specific design details and subpopulations the field would benefit from stronger engagement in open science research practices and data sharing.
AB - Real-time fMRI neurofeedback is an increasingly popular neuroimaging technique that allows an individual to gain control over his/her own brain signals, which can lead to improvements in behavior in healthy participants as well as to improvements of clinical symptoms in patient populations. However, a considerably large ratio of participants undergoing neurofeedback training do not learn to control their own brain signals and, consequently, do not benefit from neurofeedback interventions, which limits clinical efficacy of neurofeedback interventions. As neurofeedback success varies between studies and participants, it is important to identify factors that might influence neurofeedback success. Here, for the first time, we employed a big data machine learning approach to investigate the influence of 20 different design-specific (e.g. activity vs. connectivity feedback), region of interest-specific (e.g. cortical vs. subcortical) and subject-specific factors (e.g. age) on neurofeedback performance and improvement in 608 participants from 28 independent experiments. With a classification accuracy of 60% (considerably different from chance level), we identified two factors that significantly influenced neurofeedback performance: Both the inclusion of a pre-training no-feedback run before neurofeedback training and neurofeedback training of patients as compared to healthy participants were associated with better neurofeedback performance. The positive effect of pre-training no-feedback runs on neurofeedback performance might be due to the familiarization of participants with the neurofeedback setup and the mental imagery task before neurofeedback training runs. Better performance of patients as compared to healthy participants might be driven by higher motivation of patients, higher ranges for the regulation of dysfunctional brain signals, or a more extensive piloting of clinical experimental paradigms. Due to the large heterogeneity of our dataset, these findings likely generalize across neurofeedback studies, thus providing guidance for designing more efficient neurofeedback studies specifically for improving clinical neurofeedback-based interventions. To facilitate the development of data-driven recommendations for specific design details and subpopulations the field would benefit from stronger engagement in open science research practices and data sharing.
KW - Functional MRI
KW - Learning
KW - Machine learning
KW - Mega-analysis
KW - Neurofeedback
KW - Real-time fMRI
UR - http://www.scopus.com/inward/record.url?scp=85107090924&partnerID=8YFLogxK
U2 - 10.1016/j.neuroimage.2021.118207
DO - 10.1016/j.neuroimage.2021.118207
M3 - Article
C2 - 34048901
AN - SCOPUS:85107090924
SN - 1053-8119
VL - 237
JO - NeuroImage
JF - NeuroImage
M1 - 118207
ER -