Schaer_subbed.qxp 11/8/09 12:51 Page 13
Brain Development
Contribution of Structural Brain Imaging to Our
Understanding of the Cortical Development Process
Marie Schaer
1,2
and Stephan Eliez
1,3
1. Department of Psychiatry, University of Geneva School of Medicine; 2. Signal Processing Institute, Swiss Federal Institute of Technology;
3. Department of Genetic Medicine and Development, University of Geneva School of Medicine
Abstract
During the last decades, neuroimaging has become a powerful tool for studying in vivo the structural correlates associated with the burst
of acquisitions and cognitive development that occur during childhood and adolescence. This report provides a historical overview of
structural neuroimaging methods, from manually delimited structures to the most recent 3D reconstructions of the brain. With a focus on
the study of cortical morphology, we propose a schematic distinction between cortical thickness measurements as a powerful indicator of
the dynamic of brain maturation during late childhood and adolescence and cortical folding (gyrification), which provides a window on early
(foetal) brain development. In turn, we postulate that this distinction provides a framework to apprehend studies on normal brain
development as well as to better identify the processes underlying the pathogenesis of psychiatric conditions.
Keywords
Neuroimaging, magnetic resonance imaging (MRI), brain morphometry, cortical thickness, gyrification
Disclosure: This research was supported by a grant from Swiss National Research Funds to Marie Schaer (323500-111165) and Stephan Eliez (3200-063135, 3232-063134
and PP00B-102864).
Acknowledgements: The authors would like to thank Bronwyn Glaser, Marie-Christine Ottet and Martin Debbané for their ongoing collaborations.
Received: 17 November 2008 Accepted: 1 April 2009
Correspondence: Marie Schaer, Service Médico-Pédagogique, 1 rue David Dufour, Case Postale 50, 1211 Geneva 8, Switzerland. E: marie.schaer@unige.ch
As recently reviewed by Martin,
1
psychiatry and neurology, which were provide a historical overview of the methodological progresses in
for a long time completely distinct medical disciplines with separate structural neuroimaging. We further discuss how each method can
departments, practices and research interests, are becoming contribute to unravelling the neural pathways leading to the emergence
increasingly interlinked. In the 1950s, Eric Kandel described academic of cognitive delay or behavioural or psychiatric disturbances. As we
psychiatry as a “psychoanalytically-based and socially oriented exclusively focus on influences of the cortical structure, the contribution
discipline that was surprisingly unconcerned with the brain as an of diffusion tensor imaging in furthering our understanding of brain
organ of mental activity”.
2
From neuropharmacological advances in development will not be detailed here, but the interested reader can
the 1960s and 1970s to more recent progress in other disciplines of refer to recent articles by others.
9,10
neuroscience, psychiatry has increasingly been confronted by
evidence for a neural basis for mental processes and psychiatric The First Quantitative Magnetic
disturbances. A growing number of psychiatric diseases are thought to Resonance Imaging Studies
arise from disrupted neurodevelopment, including schizophrenia,
3–5
Generally speaking, neuroimaging data can be exploited on a qualitative
autism
6
or even addiction disorders.
7,8
Until recently, opportunities to basis or with quantitative measurements. The qualitative approach is
address the relationship between structural brain development and relatively similar to a clinical interpretation of the radiological exams,
cognitive performance and behavioural or psychiatric disturbances distinguishing between normal or pathological brain structure or grading
largely relied on post mortem brain examinations. This raised issues the severity or the number of specific lesions. In contrast, quantitative
such as a paucity of data, potential bias of the findings by the cause of techniques aim to measure objective values such as the volume, depth,
the death or difficulties in taking precise, quantitative measurements. length or thickness of specific cerebral structures. In such approaches,
image-processing algorithms exploit the intensity distribution of the
Nowadays, the excellent spatial resolution and soft-tissue contrast image (histogram) to distinguish between grey and white matter and
provided by magnetic resonance imaging (MRI) offer unsurpassed cerebrospinal fluid (CSF) – a process known as segmentation.
access to in vivo brain development. Being non-invasive, MRI exams Probabilistic maps reflecting the distribution of the different tissue
have become widely generalised during the last three decades in both classes are created, and overall volumes of the different
clinical and research settings. Improvements of image quality, along with tissue compartments (total grey and white matter) are computed as the
the development of computer power, allowed continued advances in sum of voxels multiplied by their tissular probability. Semi-automated
image-processing techniques. Consequently, many large-scale studies subdivisions are then typically used to obtain a more specific
have now been conducted using imaging data from typically developing localisation in the cerebral lobes and cerebellar volumes. The Talairach
individuals or patients affected by various conditions. In this article we grid is frequently used to automatically subdivide the brain into the four
© TOUCH BRIEFINGS 2009 13