% IMPORTANT: The following is UTF-8 encoded. This means that in the presence
% of non-ASCII characters, it will not work with BibTeX 0.99 or older.
% Instead, you should use an up-to-date BibTeX implementation like “bibtex8” or
% “biber”.
@ARTICLE{He:865865,
author = {He, Tong and Kong, Ru and Holmes, Avram J. and Nguyen, Minh
and Sabuncu, Mert R. and Eickhoff, Simon B. and Bzdok,
Danilo and Feng, Jiashi and Thomas Yeo, B. T.},
title = {{D}eep neural networks and kernel regression achieve
comparable accuracies for functional connectivity prediction
of behavior and demographics},
journal = {NeuroImage},
volume = {206},
issn = {1053-8119},
address = {Orlando, Fla.},
publisher = {Academic Press},
reportid = {FZJ-2019-05152},
pages = {116276},
year = {2020},
abstract = {There is significant interest in the development and
application of deep neural networks (DNNs) to neuroimaging
data. A growing literature suggests that DNNs outperform
their classical counterparts in a variety of neuroimaging
applications, yet there are few direct comparisons of
relative utility. Here, we compared the performance of three
DNN architectures and a classical machine learning algorithm
(kernel regression) in predicting individual phenotypes from
whole-brain resting-state functional connectivity (RSFC)
patterns. One of the DNNs was a generic fully-connected
feedforward neural network, while the other two DNNs were
recently published approaches specifically designed to
exploit the structure of connectome data. By using a
combined sample of almost 10,000 participants from the Human
Connectome Project (HCP) and UK Biobank, we showed that the
three DNNs and kernel regression achieved similar
performance across a wide range of behavioral and
demographic measures. Furthermore, the generic feedforward
neural network exhibited similar performance to the two
state-of-the-art connectome-specific DNNs. When predicting
fluid intelligence in the UK Biobank, performance of all
algorithms dramatically improved when sample size increased
from 100 to 1000 subjects. Improvement was smaller, but
still significant, when sample size increased from 1000 to
5000 subjects. Importantly, kernel regression was
competitive across all sample sizes. Overall, our study
suggests that kernel regression is as effective as DNNs for
RSFC-based behavioral prediction, while incurring
significantly lower computational costs. Therefore, kernel
regression might serve as a useful baseline algorithm for
future studies.},
cin = {INM-7},
ddc = {610},
cid = {I:(DE-Juel1)INM-7-20090406},
pnm = {571 - Connectivity and Activity (POF3-571)},
pid = {G:(DE-HGF)POF3-571},
typ = {PUB:(DE-HGF)16},
pubmed = {pmid:31610298},
UT = {WOS:000507987000012},
doi = {10.1016/j.neuroimage.2019.116276},
url = {https://juser.fz-juelich.de/record/865865},
}