Structural MRI Predicts Biological Maturity
-
1
Montreal Neurological Institute, Canada
-
2
Tampere University of Technology, Finland
Structural MRI has considerably advanced our knowledge about the complicated process behind brain maturation. Developmental trajectories of grey matter show an inverted-U shaped pattern peaking around puberty, suggesting a biological mechanism of synaptic overproduction followed by pruning [Giedd et al. 1999; Giedd, 2008]. Cognitive milestones happen concurrently with these structural changes, and a delay in such changes has been implicated in developmental disorders such as ADHD [Giedd and Rapoport, 2010]. Accurate estimation of individuals' brain maturity, therefore, is critical in establishing a baseline for normal brain development against which neurodevelopmental disorders can be assessed. In this study, structural MRI scans from a large dataset of normally growing children and adolescents (n = 308) with age ranging from 6 to 22 years were used [Evans, 2006]. Using a completely automated and well-validated pipeline (CIVET), morphological parameters namely cortical thickness and surface area were calculated at 81,924 vertices covering entire cortex. These high-dimensional morphological data were down-sampled to 2,560 points for each morphological parameter and used to build a predictive model to estimate the chronological age via the elastic-net regularized regression [Friedman et al. 2010] that jointly performs variable selection and model estimation. Nested 10-fold balanced cross-validation was used to estimate the model tuning parameters (the inner cross-validation loop) and the prediction error (the outer cross-validation loop). The resultant model accounted for 62% of the sample variance and had mean absolute error of 630 days in the (outer) cross-validation loop (see Figure 1 for a scatter plot where females and males are denoted with blue and green colors) while it provided nearly perfect estimates during the training (variance explained 95% and the mean absolute error of 232 days). In the model averaged over cross-validation folds, we observed that a widely distributed pattern of cortical parameters were utilized in age estimation. The results of our study demonstrate that structural MRI can be used to predict individuals' biological maturity with high accuracy, a critical information that might help in discerning individuals with neurodevelopmental disorders.
Acknowledgements
BSK is supported by a Post Doctoral Fellowship from Fonds de la recherche en santé du Québec (FRSQ), Quebec, Canada.
References
Giedd et al. 1999, "Brain development during childhood and adolescence: a longitudinal MRI study", Nature Neuroscience 2, 61-863.
Giedd 2008, "The Teen Brain: Insights from Neuroimaging", Journal of Adolescent Health 42, 335-343.
Giedd and Rapoport 2010, "Structural MRI of Pediatric Brain Development: What Have We Learned and Where Are We Going?", Neuron 67, 728-734.
Evans 2006, "The NIH MRI study of normal brain development", NeuroImage 30, 184–202.
Friedman et al. 2010, "Regularization paths for generalized linear models via coordinate descent", Journal of Statistical Software 33, 1–22.
Keywords:
Neuroimaging,
cortical thickness,
maturation,
Elastic-net Regularized Regression,
Neurodevelopmental disorders
Conference:
Neuroinformatics 2013, Stockholm, Sweden, 27 Aug - 29 Aug, 2013.
Presentation Type:
Poster
Topic:
Neuroimaging
Citation:
Khundrakpam
BS,
Tohka
J and
Evans
AC
(2013). Structural MRI Predicts Biological Maturity.
Front. Neuroinform.
Conference Abstract:
Neuroinformatics 2013.
doi: 10.3389/conf.fninf.2013.09.00104
Copyright:
The abstracts in this collection have not been subject to any Frontiers peer review or checks, and are not endorsed by Frontiers.
They are made available through the Frontiers publishing platform as a service to conference organizers and presenters.
The copyright in the individual abstracts is owned by the author of each abstract or his/her employer unless otherwise stated.
Each abstract, as well as the collection of abstracts, are published under a Creative Commons CC-BY 4.0 (attribution) licence (https://creativecommons.org/licenses/by/4.0/) and may thus be reproduced, translated, adapted and be the subject of derivative works provided the authors and Frontiers are attributed.
For Frontiers’ terms and conditions please see https://www.frontiersin.org/legal/terms-and-conditions.
Received:
29 Apr 2013;
Published Online:
11 Jul 2013.
*
Correspondence:
Dr. Budhachandra S Khundrakpam, Montreal Neurological Institute, Montreal, Quebec, H3A 2B4, Canada, budhakh@gmail.com