TY - GEN
T1 - Sex Differences in Predicting Fluid Intelligence of Adolescent Brain from T1-Weighted MRIs
AU - Ranjbar, Sara
AU - Singleton, Kyle W.
AU - Curtin, Lee
AU - Massey, Susan Christine
AU - Hawkins-Daarud, Andrea
AU - Jackson, Pamela R.
AU - Swanson, Kristin R.
N1 - Funding Information:
The authors would like to thank the Challenge Organizers and ABCD Study Researchers for the opportunity to participate and utilize their data. We also thank Kevin Flores, Erica Rutter, and John Nardini for many helpful discussions. Further, we acknowledge the following funding sources: James S. McDonnell Foundation, U54CA210180, U54CA193489, 3U54CA193489-04S3, and U01CA220378.
Publisher Copyright:
© 2019, Springer Nature Switzerland AG.
PY - 2019
Y1 - 2019
N2 - Fluid intelligence (Gf) has been defined as the ability to reason and solve previously unseen problems. Links to Gf have been found in magnetic resonance imaging (MRI) sequences such as functional MRI and diffusion tensor imaging. As part of the Adolescent Brain Cognitive Development Neurocognitive Prediction Challenge 2019, we sought to predict Gf in children aged 9–10 from T1-weighted (T1 W) MRIs. The data included atlas–aligned volumetric T1 W images, atlas–defined segmented regions, age, and sex for 3739 subjects used for training and internal validation and 415 subjects used for external validation. We trained sex-specific convolutional neural net (CNN) and random forest models to predict Gf. For the convolutional model, skull-stripped volumetric T1 W images aligned to the SRI24 brain atlas were used for training. Volumes of segmented atlas regions along with each subject’s age were used to train the random forest regressor models. Performance was measured using the mean squared error (MSE) of the predictions. Random forest models achieved lower MSEs than CNNs. Further, the external validation data had a better MSE for females than males (60.68 vs. 80.74), with a combined MSE of 70.83. Our results suggest that predictive models of Gf from volumetric T1 W MRI features alone may perform better when trained separately on male and female data. However, the performance of our models indicates that more information is necessary beyond the available data to make accurate predictions of Gf.
AB - Fluid intelligence (Gf) has been defined as the ability to reason and solve previously unseen problems. Links to Gf have been found in magnetic resonance imaging (MRI) sequences such as functional MRI and diffusion tensor imaging. As part of the Adolescent Brain Cognitive Development Neurocognitive Prediction Challenge 2019, we sought to predict Gf in children aged 9–10 from T1-weighted (T1 W) MRIs. The data included atlas–aligned volumetric T1 W images, atlas–defined segmented regions, age, and sex for 3739 subjects used for training and internal validation and 415 subjects used for external validation. We trained sex-specific convolutional neural net (CNN) and random forest models to predict Gf. For the convolutional model, skull-stripped volumetric T1 W images aligned to the SRI24 brain atlas were used for training. Volumes of segmented atlas regions along with each subject’s age were used to train the random forest regressor models. Performance was measured using the mean squared error (MSE) of the predictions. Random forest models achieved lower MSEs than CNNs. Further, the external validation data had a better MSE for females than males (60.68 vs. 80.74), with a combined MSE of 70.83. Our results suggest that predictive models of Gf from volumetric T1 W MRI features alone may perform better when trained separately on male and female data. However, the performance of our models indicates that more information is necessary beyond the available data to make accurate predictions of Gf.
KW - Deep learning
KW - Fluid intelligence
KW - Sex differences
UR - http://www.scopus.com/inward/record.url?scp=85075658459&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85075658459&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-31901-4_18
DO - 10.1007/978-3-030-31901-4_18
M3 - Conference contribution
AN - SCOPUS:85075658459
SN - 9783030319007
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 150
EP - 157
BT - Adolescent Brain Cognitive Development Neurocognitive Prediction - 1st Challenge, ABCD-NP 2019, held in Conjunction with MICCAI 2019, Proceedings
A2 - Pohl, Kilian M.
A2 - Adeli, Ehsan
A2 - Thompson, Wesley K.
A2 - Linguraru, Marius George
PB - Springer
T2 - 1st Challenge in Adolescent Brain Cognitive Development Neurocognitive Prediction, ABCD-NP 2019, held in conjunction with the 22nd International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2019
Y2 - 13 October 2019 through 13 October 2019
ER -