Logo Logo

Gaus, Richard; Pölsterl, Sebastian; Greimel, Ellen; Schulte‐Körne, Gerd; Wachinger, Christian (2023): Can we diagnose mental disorders in children? A large‐scale assessment of machine learning on structural neuroimaging of 6916 children in the adolescent brain cognitive development study. JCPP Advances. ISSN 2692-9384

[thumbnail of JCPP Advances - 2023 - Gaus - Can we diagnose mental disorders in children  A large‐scale assessment of machine learning on.pdf] Published Article
JCPP Advances - 2023 - Gaus - Can we diagnose mental disorders in children A large‐scale assessment of machine learning on.pdf

The publication is available under the license Creative Commons Attribution.

Download (913kB)

Abstract

Background
Prediction of mental disorders based on neuroimaging is an emerging area of research with promising first results in adults. However, research on the unique demographic of children is underrepresented and it is doubtful whether findings obtained on adults can be transferred to children.

Methods
Using data from 6916 children aged 9–10 in the multicenter Adolescent Brain Cognitive Development study, we extracted 136 regional volume and thickness measures from structural magnetic resonance images to rigorously evaluate the capabilities of machine learning to predict 10 different psychiatric disorders: major depressive disorder, bipolar disorder (BD), psychotic symptoms, attention deficit hyperactivity disorder (ADHD), oppositional defiant disorder, conduct disorder, post-traumatic stress disorder, obsessive-compulsive disorder, generalized anxiety disorder, and social anxiety disorder. For each disorder, we performed cross-validation and assessed whether models discovered a true pattern in the data via permutation testing.

Results
Two of 10 disorders can be detected with statistical significance when using advanced models that (i) allow for non-linear relationships between neuroanatomy and disorder, (ii) model interdependencies between disorders, and (iii) avoid confounding due to sociodemographic factors: ADHD (AUROC = 0.567, p = 0.002) and BD (AUROC = 0.551, p = 0.002). In contrast, traditional models perform consistently worse and predict only ADHD with statistical significance (AUROC = 0.529, p = 0.002).

Conclusion
While the modest absolute classification performance does not warrant application in the clinic, our results provide empirical evidence that embracing and explicitly accounting for the complexities of mental disorders via advanced machine learning models can discover patterns that would remain hidden with traditional models.
Prediction of mental disorders based on neuroimaging is an emerging area of research with promising first results in adults. However, research on the unique demographic of children is underrepresented and it is doubtful whether findings obtained on adults can be transferred to children.
Methods

Using data from 6916 children aged 9–10 in the multicenter Adolescent Brain Cognitive Development study, we extracted 136 regional volume and thickness measures from structural magnetic resonance images to rigorously evaluate the capabilities of machine learning to predict 10 different psychiatric disorders: major depressive disorder, bipolar disorder (BD), psychotic symptoms, attention deficit hyperactivity disorder (ADHD), oppositional defiant disorder, conduct disorder, post‐traumatic stress disorder, obsessive‐compulsive disorder, generalized anxiety disorder, and social anxiety disorder. For each disorder, we performed cross‐validation and assessed whether models discovered a true pattern in the data via permutation testing.
Results

Two of 10 disorders can be detected with statistical significance when using advanced models that (i) allow for non‐linear relationships between neuroanatomy and disorder, (ii) model interdependencies between disorders, and (iii) avoid confounding due to sociodemographic factors: ADHD (AUROC = 0.567, p = 0.002) and BD (AUROC = 0.551, p = 0.002). In contrast, traditional models perform consistently worse and predict only ADHD with statistical significance (AUROC = 0.529, p = 0.002).
Conclusion

While the modest absolute classification performance does not warrant application in the clinic, our results provide empirical evidence that embracing and explicitly accounting for the complexities of mental disorders via advanced machine learning models can discover patterns that would remain hidden with traditional models.
06 28 2023 e12184 10.1002/jcv2.12184 Bundesministerium für Bildung und Forschung https://doi.org/10.13039/501100002347 031L0200A http://creativecommons.org/licenses/by/4.0/ 10.1002/jcv2.12184 https://acamh.onlinelibrary.wiley.com/doi/10.1002/jcv2.12184 10.1016/j.neuroimage.2020.117002 10.1097/00004583-200001000-00016 10.1176/appi.books.9780890425596 10.1016/j.neuroimage.2016.02.079 10.1111/jcpp.13481 10.1016/j.dcn.2017.10.010 10.1016/j.dcn.2021.101031 10.1080/15374416.2020.1750022 10.1097/00004583-199201000-00012 10.1017/S0140525X17002266 10.1109/ICPR.2010.764 10.1016/j.bpsc.2017.11.007 10.1016/j.dcn.2018.03.001 10.1001/jamapediatrics.2019.2081 10.1037/a0038498 10.1176/appi.ajp.2018.1750701 10.1016/j.neuroimage.2012.01.021 10.2307/2699986 10.1016/j.dcn.2018.04.004 10.1001/jamapsychiatry.2014.2206 10.1148/radiology.143.1.7063747 10.1101/2020.02.10.942011 10.3389/fpsyt.2018.00242 10.1016/j.biopsych.2016.10.028 10.1038/mp.2012.105 10.1038/s41386-020-0736-6 10.1097/00004583-199707000-00021 10.1016/j.neuroimage.2022.119046 10.1016/j.jpsychires.2020.11.023 10.1176/ajp.151.8.1163 10.1201/9781315382715 10.1111/jcpp.12611 10.1037/0021-843x.107.2.305 10.1016/j.bpsc.2019.11.007 10.1109/ICDM.2009.108 10.1038/s41398-020-01192-8 10.1007/bf00927116 Pohl K. M. Thompson W. K. Adeli E. Landman B. A. Linguraru M. G. &Tapert S. F.(2021).Adolescent brain cognitive development neurocognitive prediction challenge. Retrieved May 11 2021 fromhttps://sibis.sri.com/abcd-np-challenge/ 10.1007/978-3-030-31901-4 10.1002/hbm.25013 10.1007/s10994-011-5256-5 10.1016/j.socscimed.2013.04.026 10.1002/hbm.20906 10.1007/s11604-018-0794-4 10.1016/j.neuroimage.2019.02.057 10.1016/j.schres.2017.10.023 10.3389/fpsyt.2016.00050 Snoek J. Larochelle H. &Adams R. P.(2012).Practical Bayesian optimization of machine learning algorithms. 10.1016/j.neuroimage.2018.09.074 10.1016/j.biopsych.2020.02.016 10.1523/jneurosci.3302-16.2017 10.1016/j.dcn.2018.12.004 10.1016/j.jaac.2019.05.009 10.1016/j.biopsych.2018.04.023 10.1016/j.media.2020.101879 10.1016/j.neubiorev.2015.08.001 10.1038/nn.4478 10.1016/j.jpsychires.2015.01.015 10.1002/1097-0142(1950)3:1<32::aid-cncr2820030106>3.0.co;2-3 10.1016/j.eswa.2017.04.003 10.1038/s41398-021-01201-4

View Item
View Item