MICCAI
2008 WORKSHOP |
Organizers:
Guido Gerig (
John H. Gilmore (UNC)
Alan Evans (MNI)
Daniel Rueckert (
Simon Warfield
(Children’s Hospital Boston)
Rationale:
Imaging studies of early brain development get
increasing attention as improved modeling of the pattern of normal development
and of change from normal might lead to a better understanding of origin,
timing and nature of morphologic and functional differences in
neurodevelopmental disorders. Measuring the trajectory of growth via
noninvasive imaging such as structural MRI and DTI, for example, will likely
provide a vastly improved understanding of early brain development, changes due
to delayed development or pathology, and its relationship to neuropsychiatric
disorders.
Studying
the age group from birth to 5 years, even including premature born babies,
involves several major challenges which are specific to neuroimaging of this
age group. Most important are the imaging of non-sedated infants with low
failure rate, in particular in longitudinal studies, and the development of appropriate
image analysis methodologies that can cope with low contrast-to-noise ratio,
rapid change of size of brain structures, complex brightness changes in MRI reflecting rapid white matter
structuring through myelination and axon elimination, rapid change and large
variability of anatomical shapes, and locally varying contrast associated with
early structuring. The study of growth trajectories by definition involves
longitudinal imaging and will require application of computational tools for
processing of 4D data (3D plus time) and statistical methods for longitudinal
data analysis.
Goals of
the workshop are the following:
·
to introduce the topics of early brain
development, pediatric imaging and its challenges, and association
of structures and patterns observed in images to what is known about the
underlying neurobiology,
·
to introduce the
clinical need and potential impact for studies of early brain development,
including the need for normative data to model trajectories of normal brain
development, the motivation for early diagnosis of children at risk for mental
illness and neurological disorders, the ability to study rare diseases,
monitoring treatment and intervention, and image modalities appropriate for
this age group,
·
to inform about publicly
available image databases of pediatric imaging studies to test, validate
and develop advanced computational tools,
·
to discuss the
current status of image analysis methodology and emerging new approaches
to study early brain growth and change from normal (including segmentation,
registration, atlas-building, computational anatomy tools, analysis of MRI,
DTI, ultrasound, and more),
·
to discuss validation
and cross-method comparison of image analysis,
·
and finally to
do brainstorming on key issues critical to further advance the field.
Format:
The workshop includes
invited talks by representatives from different institutions strongly involved
in pediatric neuroimaging studies and analysis of such data and also short
poster teaser presentations and posters of submitted abstracts. Participants
are encouraged to use the discussion time after presentations and the
brainstorming and panel sessions for a critical dialogue to clarify the state-of-the-art
and formulate outstanding issues both w.r.t. imaging technology, image analysis
challenges, and fundamental mathematical and algorithmic issues related to this
type of longitudinal data presenting rapid change of contrast, size, shapes and
appearance. The workshop will also inform about availability of image
databases, pediatric atlases, and tools.
Participants will get pdf
copies of talks and presentations (ready on CD at workshop).
Intended audience:
Researchers from medicine,
biology, imaging, bioengineering, image analysis, and biostatistics, among
others, who are interested or already involved in studies of early brain
development in healthy and disease, and also researchers who are developing
imaging/image analysis technologies which are of particular interest to study
this type of image data.