Structural Imaging and Neuroanatomy Research
Unit
DIRECTOR: NANCY C. ANDREASEN, M.D.,
PH.D.
CO-DIRECTORS: MICHAEL VANNIER, M.D., JAMES EHRHARDT, Ph.D.
Overall Aims
1. To use MR to identify and study the neuroanatomic abnormalities
that occur in disease states, such as schizophrenia or autism.
2. To use MR to examine neuroanatomic measures that may
provide information about neurodevelopmental processes, such as gyrification
patterns.
3. To use MR to study and measure specific subregions
hypothesized to be involved in dysfunctional anatomic/functional circuitry
in schizophrenia, such as prefrontal cortex, thalamus, and cerebellum.
4. To use MR to examine anatomic circuitry directly through
the use of diffusion anisotropy.
5. To use MR to study factors that could confound measurements
of brain structure in schizophrenia, such as hydrational status.
6. To identify and develop new methods that will facilitate
these aims, such as novel scanning sequences or new approaches to image
analysis (e.g., neural nets, global pattern matching).
7. To confirm MR findings and improve their spatial resolution through
the study of postmortem tissue.
Background and Rationale
The development of the technology of in vivo brain
imaging during the past 15 years has provided an extraordinarily powerful
group of tools for understanding the human brain. These tools permit us
to study brain anatomy in exquisite detail, to observe shifts in metabolic
activity as the brain responds to cognitive or perceptual challenges, and
to quantitatively measure the neurochemical activity of a variety of neurotransmitter
systems. Prior to the development of in vivo brain imaging, investigators
were limited to indirect inferential techniques such as cognitive tests,
the study of peripheral metabolites, or the examination of postmortem tissue.
These earlier techniques imposed substantial technical limitations on investigations,
making it difficult to disentangle the effects of aging and medication
or interactions between the peripheral and central nervous system. In
vivo brain imaging techniques allow us to study the human brain directly
in both normal individuals and in patients suffering from major mental
illnesses. These techniques offer the possibility of being able to identify
the anatomic and physiologic substrates of these illnesses.
Our own work and that of others has repeatedly demonstrated that patients with schizophrenia have structural brain abnormalities (e.g., Johnstone et al 1976; Weinberger et al 1979, 1980b, 1988; Andreasen et al 1982d, 1986d, 1990b, 1994i; and many, many more). The challenge currently facing investigators in psychiatry is to build on the burgeoning knowledge base in basic neuroscience and the explosion in imaging and computer science technology in order to choose from a large array of hypotheses and techniques the ones that are most likely to advance our understanding of the normal brain and of major mental illnesses.
Proposed MR studies involve the examination of a series of scientific questions in cognitive neuroscience, systems level and developmental neurobiology, and clinical psychopathology. The studies focus on regions hypothesized to be important parts of the neural circuitry used in the cognitive processes that are dysfunctional in schizophrenia and other developmental disorders with shared phenomenology such as autism. Methods and regions have been chosen for study because they may elucidate this circuitry and the developmental processes that shape it from fetal life through young adulthood. Neurodevelopmental abnormalities have been an important working hypothesis for the mechanisms that produce schizophenia for many years (Feinberg, 1982, Andreasen et al 1986a, 1986d; Weinberger, 1987; Swayze et al 1990). Our more recent work, using both MR and PET, has led us to postulate that abnormal neurodevelopmental mechanisms may preferentially affect the connectivity between midline regions, such as the thalamus, and cortical regions, leading to impairment in the ability to fluidly coordinate mental activities (Andreasen et al 1996k, 1997g, in press a).
The projects in this research unit are designed to illuminate various facets of this hypothesis. Seven projects are included. The first project continues our ongoing work to examine the neuroanatomic substrates of schizophrenia using MR, with an emphasis on examining specific nodes on the CCTCC. The second project tackles the difficult problem of measuring thalamic subnuclei in detail: developing a new scanning sequence and examining the reliability and validity of this new approach to visualization and measurement. The third project extends the investigation of the thalamus in schizophrenia to the cellular and molecular level, through a study that will use postmortem tissue to examine region-specific alterations in cell number and neurotransmitter distribution in the dorsal thalamus. The fourth project uses MR to examine a childhood-onset neurodevelopmental disorder, autism, employing measurements similar to those used in our adult schizophrenic sample, in order to identify patterns of anomalous organization and connectivity. The fifth project proposes to use cutting edge MR image analysis techniques in order to examine brain development in children and adolescents and to study individual variability due to gender effects, handedness, or laterality. The sixth project employs a specific new MR technique, anisotropic diffusion imaging, to directly examine the connectivity of brain regions in vivo and compare prefrontal-cerebellar-thalamic circuitry in patients with schizophrenia and normal controls. The seventh project conducts experimental studies designed to examine one potential confounder of structural MR measurements, hydration status, within the context of several studies that we have already conducted to date that indicate that anatomic measures are not static and that they may be changed by exogenous factors such as exposure to neuroleptic medication (Rodriguez et al submitted; Swayze et al 1992; Westmoreland et al submitted).
Progress Report
This research unit has made substantial progress over
the past four years. While external grant support is only an indirect indicator
of achievement, nonetheless it is an important one. In our last application,
only one project in the MR Research Unit (as this unit was then named)
had external support. In this application, five of the seven projects have
external funding. Two others are promising pilot projects for which we
request seed money support so that external funding can be obtained. More
importantly, however, this unit has also made substantive scientific progress.
In the previous review, reviewers were concerned that the proposed studies
lacked a unifying and innovative group of hypotheses to be tested. Our
work completed during the past four years--combining image analysis techniques
with MR and PET findings--has permitted us to clearly articulate a major
direction and some heuristic hypotheses that we believe likely to advance
knowledge about the mechanisms of schizophrenia. Our MR studies have provided
an important group of findings that support conceptualizing schizophrenia
as a neurodevelopmental disorder that is due to dysfunctional neural circuits
(and specifically the CCTCC). Further, some of our results suggest that
the abnormality may be structural in addition to metabolic/physiologic/neurochemical.
The studies proposed for the coming five years follow up on these findings
in various ways. One of our two pilot studies uses the substantial expertise
of Martin Cassell to examine the neuropathology of thalamic subnuclei (Project
3), while the other extends MR technology to in vivo study of neuronal
connectivity through anisotropic diffusion imaging (Project 6). Other studies
propose to use structural MR to examine nodes on the CCTCC and to use gyrification
measures to examine brain development (Projects 1 and 2), to examine structural
abnormalities in a related neurodevelopmental disorder, autism (Project
4), to study neurodevelopment and gender differences in normals (Project
5), and to examine the effects of hydration on measurement of brain structure
(Project 7).
Much of the progress achieved to date has depended
on the continuing development of new and powerful methods for simplifying
and automating the process of measuring and analyzing MR data, through
the further development of our software package, BRAINS. Some of this
progress is summarized in the Image Analysis Core Unit, which outlines
the various tools available in BRAINS and describes our work to date
in assessing their reliability and validity. Briefly, our progress
with image analysis includes the development of:
Volume rendering software that permits simultaneous visualization
of anatomy in three orthogonal planes and visualization of brain surface
anatomy, tracing on either brain surface or internal structures, telegraphing
of the coordinates across the other portions of the image to produce very
accurate measurement of structures based on simultaneous three-dimensional
visualization, and calculation of areas and volumes of the structures/regions
traced.
Standard and reliable methods for manually tracing individual
brain structures or regions, such as lobes of the brain, the hippocampus,
or the caudate nucleus.
Automated atlas-based methods for subdividing the brain
into specific lobes.
Both linear and nonlinear methods for normalizing brains
so that subjects can be averaged and so that both subcortical and gyral
anatomy can be made as comparable as possible across individuals.
Methods to measure differences in the shape of brain subregions
(similar to the thin plate spline technique).
Multispectral methods for achieving precise segmentation
of brain tissue into gray matter (GM), white matter (WM), and cerebrospinal
fluid (CSF).
Image subtraction techniques as a tool for doing cross-group
comparisons of differences.
Methods for visualizing and measuring sulcal and gyral
anatomy, including validation.
These various methods, which are the consequence of
many years of effort in developing well-validated and accurate methods
for measuring brain anatomy, are nearly all completely automated. Having
automated methods produces many advantages. It permits the study of
large samples with relative speed and efficiency. It eliminates laborious
hand-tracing, which is not only time-consuming, but also requires careful
calibration across tracers and can potentially be less reliable if
rater drift occurs.
The techniques to measure surface anatomy were particularly difficult
to achieve, since they required that we solve some of the fundamental problems
in finding an accurate brain surface, such as "touching gyri" and "buried
cortex." We now have, however, a valid and sensitive group of measures
of gyrification, including a classic Zilles gyrification index, a 3D adaptation
of it, a measure of brain surface area, a measure of cortical thickness
for the entire surface, and measures of the degree of sulcal and gyral
curvature. These measures will assist us in using indices of gyrification
as a tool to probe neurodevelopmental abnormalities.
Progress with the Study of Anatomic/Circuit
Abnormalities in Schizophrenia Using MR
Our progress in this area includes finding:
These findings provide convergent evidence from multiple
structural imaging studies that patients with schizophrenia have measurable
abnormalities in the nodes in the CCTCC. Nonetheless, methods used
to date are relatively crude, since we have only measured large regions
or structures that are known to have important anatomical and functional
subdivisions. Consequently, our future structural MR studies of schizophrenia
will focus on the development and application of more fine-grained
methods for conducting anatomic measurements of the prefrontal cortex,
the cerebellum, and the thalamus.
© The University of Iowa 2005. All rights reserved.
Latest update May 7, 2005 Webpages maintained by Hans J. Johnson. E-mail the webmaster
MB