| |
|
|
| |
The research in
the unit is based on the combined application
of various mapping techniques for in-vivo study
of the human brain structure and function. Currently
the main methodology of the unit is based on advanced
Magnetic Resonance Imaging (MRI). However, the
unit aim is to combine various methods that are
available nowadays for functional mapping of the
human brain. |
|
| |
The following describes
the main research areas that have already been
pursued in the joint unit by applying advanced
MRI techniques in humans: |
|
| |
A. Cognitive
Neuroscience |
|
| |
Asking questions regarding
"interface processes" in the brain (visual-cognitive,
visual-emotion, emotional-cognitive), language
processing and representation in the brain (computational
aspects in syntax, semantic components and mathematical
faculties), and the effect of interpersonal interaction
emotional experience on the brain. Examples of
current projects: |
|
| |
Object recognition
- Mapping of high order visual areas in relation
to stimuli category stimuli type and task. Functional
organization and the effect of early visual experience
(in collaboration with the Dr Malach Weizmann
institute). Brain reorganization following abnormal
development of the visual cortex is studied in
Amblyopia patients. |
|
| |
Language - Mapping
the representation of syntactic and semantic principle
of language functions. |
|
| |
Emotion in the brain
- Mapping the differential activation in limbic
sensory and cognitive area for valence and arousal
dimensions of emotion. Studying the longitudinal
effects of psychological stress |
|
| |
|
|
| |
Studying the dynamic
of changes (plasticity) in brain tissue (gray
and white matter) and its functionality that allow
for normal development and re-habilitation following
insult. Examples of projects: |
|
| |
The developing brain:
Gray and white matter differentiation from
fetus to adulthood, developmental disorders (autism,
Down's Syndrome and others). |
|
| |
The aging brain:
Functional and structural brain changes in
normal aging, mild cognitive impairment and dementia
(vascular dementia, Alzheimer and CJD) |
|
| |
The insulted brain
and its rehabilitation: Functional and structural
brain plasticity following focal vascular lesions. |
|
| |
The emotionally
disturbed brain: Visual-limbic effects of
traumatic experience (Post Traumatic Stress Disorder),
brain representation of face perception and interpretation
in psychotics (schizophrenia), processing of emotionally
ambivalent stimuli in the psychogenic disorder,
affective versus somatic aspects of pain disorder.
|
|
| |
C. Brain
Mapping Methodology |
|
| |
Currently we are focusing
on the development of methods for dynamic imaging
and improved spatial and temporal resolution in
fMRI and diffusion tensor imaging. Methods such
as keyhole imaging, k-space minimization algorithms
together with advanced image processing constitute
our main areas of investment. High b-value diffusion
tensor imaging is a new technique we are investigating,
with exciting possibilities in the field of white
matter disease. Analytical and numerical models
of tissue structure assist us in interpreting
the imaging data. In the future we will apply
advanced statistical methods as well as mathematical
models of brain functions to the analyzed brain
mapping data. Example of current projects: |
|
| |
Improving MRI methodology
: higher resolution MRI; 3D presentation of
gray and white matter, new methods for specific
tissue probing. |
|
| |
Combined methodologies:
fMRI-DTI, fMRI, ERPs, MR Spectroscopy-DTI and
fMRI. |
|
| |
New advanced applications
in MRI: high b-value of Diffusion MRI and
MR-spectroscopy (white matter and gray matter
degeneration). |
|
| |
|
|
| |
|
|
| |
A. Cognitive
Neuroscience |
|
| |
1. Object selective organization
in visual cortex: |
|
| |
From normal visual
experience to cortical plasticity due to reduced
vision. |
|
| |
|
|
| |
This work revealed
an association between object images and the organization
of visual field eccentricity. We have shown that
brain regions representing object categories that
rely on detailed central scrutiny such as faces
are more strongly associated with processing of
central information compared with representation
of objects that may be recognized by more peripheral
information such as buildings (Nature Neuroscience-Amedi
et al, 4(3): 324-330, 2001; Cerebral Cortex -
Lerner et al , 12: 163-177, 2002 and 11: 287-297,
2001; Neuron - Hasson et al. 34: 479-490,
2002). Lately, we also demonstrated that the central
biased regions are more sensitive to letters and
words than the periphery biased areas. The above
projects are being done in collaboration with
the research group headed by Dr Rafael Malach,
Neurobiology Department Weizmann Institute. |
|
| |
Object Organization in
Amblyopia
|
|
| |
|
|
| |
In an ongoing study
on patients with Amblyopia ("lazy eye") we have
found a disturbed organization in this object
specific area. In the amblyopic eye that experienced
less central vision throughout life, the face-related
central representation in the cortex was shrunk
relative to the sound eye in the same patient
(see below marked in red circle a flattened surface
of visual demonstrating this effect in an amblyopic
patient versus healthy control) |
|
| |
2. Emotion - Perception Interplay
in the Brain. |
|
| |
Differential
sensitivity of visual and emotional brain areas
to affective stimuli: From healthy to abnormal
perceptual experience. |
|
| |
For some time, researchers
have been trying to determine whether the emotional
attributes of a stimulus affect the way we perceive
it and process it. Do we see the features of a
bizarre face first and then acknowledge its oddity
or do both processes somehow occur simultaneously?
Does the bizarreness of the face influence how
we perceive the facial features to begin with?
To address these questions, it would be necessary
to analyze the response of our brain to identical
visual stimuli with varying emotional attributes,
such as unpleasantness or bizarreness. Since it
is not trivial to change the emotional load of
a stimulus without altering its visual appearance
dramatically, a definitive answer to these questions
has remained elusive. We have devised an experimental
protocol that overcomes this problem. The findings,
reported in Neuron 32: 747-757, 2001 (Rotshtein
et al, see cover) shed some light on the interaction
between emotion and visual perception. Classical
visual illusion was used in the experiment: by
simply inverting the eyes and the mouth in a face
one can dramatically change its emotional impact.
The resulting face looks bizarre but its visual
features remain largely unchanged (see picture).
The results suggest that the emotional load of
a stimulus does not affect the way we visually
respond but does have an effect on how we become
used to it if we see it many times. Imaging of
the brains of subjects looking at these faces
showed that the amygdala, a brain region implicated
in emotional response, might play a role in affecting
the visual cortex response to this emotion-dependent
adaptation.
|
|
| |
|
|
| |
|
|
Amaral et al, 1992
|
|
|
| |
White Matter Mapping
with MRI-DTI of the connection between emotional
- perceptual brain areas (marked in blue). |
|
| |
3. Emotion interacts with Decision
Process in the Brain |
|
| |
Could brain activity
reveal a covert choice? |
|
| |
Decision process under
uncertainty and risk is accompanied by distinct
emotion. It was assumed that brain activity in
the amygdala differentiates between risky and
safe choices. In order to test this hypothesis
we monitored brain activity while subjects were
engaged in a gambling game against the experimenter.
In order to win players had to occasionally choose
to bluff their opponent and risk getting caught
and suffer a loss. Following such a choice the
brain responded with increased activation while
subject was waiting to see if his choice would
be exposed or not. Thus, the amygdala response
revealed subject's covert choice based on its
moral and emotional tagging (Kahn et al Neuron,
33: 983-994, 2002). |
|
| |
|
Frontal Lobe Activation
during risk taking
|
Amygdale response during anticipation
following a risky choice |
|
|
| |
|
|
| |
|
|
| |
1.
Effect of traumatic emotional experience on the
brain |
|
| |
Several important questions
relate to this topic of our research: Does
traumatic experience affect the way we perceive
things in the world? Will this effect be detected
already in sensory areas of the brain? Does trauma
effect in visual cortex interact with awareness
and consciousness? Does trauma effect differentiate
between sensory and emotional brain regions? Findings
of this work were published in the Economist (June
23rd 2001, see below). It was also presented at
scientific meetings, and in Cellular & Molecular
Neurobiology , 21(6): 733-752, 2002 (Hendler
et al ) |
|
| |
|
|
| |
In this study we explored
brain responses to traumatic reminders in combat
veterans with and without posttraumatic stress
disorder (PTSD). We showed an increased sensitivity
of the visual cortex to combat content in pictures
only when they were presented below recognition
threshold or when presented repeatedly. In contrast,
Amygdala in PTSD was hyperactive to both combat
and neutral content in all presentation conditions.
These findings suggest that already at the sensory
cortex trauma affects the threshold of response
in the brain even beyond awareness. |
|
| |
|
|
| |
|
|
| |
2. Differential
Probing of Brain Tissue: |
|
| |
Applications in the healthy
and diseased brain |
|
| |
| To complement
our functional brain mapping our team has
implemented the high b value Diffusion Weighted
Imaging methodology to delineate the brain's
white matter (see below in methodology studies).
In the healthy brain - we focused on neuronal
maturation. Conventional MR imaging can
detect white matter maturation up to 2 years
of age. With this new technique we can now
show white matter maturation up to 6 years
of age with more minor changes up to age
22. |
 |
|
|
| |
White Matter Diseases:
|
Multiple Sclerosis
is an autoimmune disease of the central
nervous system characterized by degeneration
of the white matter. Conventional MR imaging
is used to detect white matter lesions.
High b- value DWI is more sensitive to white
matter integrity than conventional MR. This
method enables early diagnosis of pathological
areas of white matter thus providing essential
information on the progression of the disease.
Images A and B show the normal white
matter (A) and gray matter (B). In Image
D areas of pathological lesions (white)
are detected. Image C shows a large central
void-in the area of disrupted white matter.
|
|
|
| |
Dementia
Alzheimer's disease (AD)
and vascular dementia (VaD) are the most common
types of dementia. The distinction between
these two entities can be assisted by high
b-value DWI. Alzheimer's is mainly a gray
matter disease and so the white matter imaging
is almost normal. In vascular dementia, massive
white matter abnormalities can be detected
by this methodology (shown on the right).
|
|
| |
|
|
| |
White Matter
Analysis:
|
|
| |
Applications in the healthy
and diseased brain
|
|
| |
MRI enable following
white matter architecture and integrity using
diffusion weighted MRI techniques. The diffusion
MRI techniques can be divided into three levels.
The first level is the basic diffusion weighted
imaging technique and the production of apparent
diffusion coefficient (ADC) image maps. The ADC
maps are widely used today in the clinics, mainly
for characterization of ischemic brain infarcts
and some kinds of brain tumors. The second level
is the diffusion tensor imaging (DTI) technique,
which enable following white matter architecture
and for some degree also white matter integrity.
The DTI techniques find the principal diffusivities
in each pixel which represent, in case of white
matter, the directionality of white matter fibers.
The third level is the high b value, q-space analyzed,
diffusion technique which is expansion of the
DTI method. This method enhances the contribution
of intra-axonal water molecules diffusion and
hence enhances the ability to detect changes in
the integrity of the axon. In our lab the diffusion
techniques is investigated under the following
topics:
1) Integration of white matter architecture and
functional activity deduced from DTI and fMRI,
respectively.
a) In healthy tissue.
b) In and surrounding brain tumors.
2) Evaluating the diagnostic ability of the high
b value, q-space analyzed, diffusion imaging techniques.
a) In white matter degeneration diseases such
as multiple sclerosis.
b) For evaluation the contribution of white matter
changes in different types of dementia (AD, VaD
and MCI).
c) Following normal myelination from infancy to
adulthood
Examples:
1) High b value, q-space analyzed diffusion
imaging in Multiple Sclerosis (MS).
MS is an autoimmune disease of the central nervous
system affecting millions worldwide. The conventional
T2 weighted MRI detects focal brain lesions in
MS diseased brains. Despite the ability to detect
lesions in MS brains, poor correlation was found
between the clinical state of the patient and
the disease load as expressed by the number and
the volume of lesions detected by T2 weighted
MRI. Therefore, it is believed that MS affects
brain areas that surpass the lesions detected
by T2 weighted MRI suggesting that the normal
appearing white matter in T2 weighted MRI might
be abnormal in those cases. This was corroborated
by 1H spectroscopic studies in MS, which detected
reduction in NAA levels in areas of normal appearing
white matter (NAWM). Diffusion tensor imaging
(DTI) shows significant reduction in the anisotropy
in areas of MS lesions. The reduction in anisotropy
was attributed to the disappearance of myelin,
which causes the diffusion to be less restricted
perpendicular to the nerve fibers leading to overall
reduced anisotropy. It is possible to detect reduction
in diffusion anisotropy in NAWM of MS patients
but only when measured on large number of subjects.
This might be related to the relatively large
contribution of extra-cellular water to the observed
signal in DTI (performed at low b values). Therefore,
for initial demyelination process, the differences
in DTI might be small and large data sets are
generally needed to detect the reduction in anisotropy.
High b value q-space analyzed DWI, which emphasize
the contribution of the intra-axonal water component,
detects abnormal tissue in NAWM with better sensitivity
than DTI. Example for q-space images of a healthy
volunteer and an MS patient is given in Figures
1 and 2, respectively. The q-space profiles of
water in MS lesions approach that of free water.
In the NAWM, it is possible to detect a restricted
narrow displacement component, however, this component
is of much lower magnitude than in the respective
areas in healthy subjects. Overall, an increase
in the apparent displacement and a decrease in
the probability for zero displacement is observed
in the lesions and in some areas of the NAWM.
The q-space probability value was found to be
in good correlation (r=0.60) with NAA/Cr ratios
obtained from 2-dimensional chemical shift imaging
(CSI) in the MS patients. Moreover, it was found
that areas that had normal NAA/Cr ratios in the
MS patient had also normal displacement and probability
values. This correlation between the q-space parameters
and NAA/Cr ration suggests that the high b value
q-space images might serve as neuronal marker
similar to NAA but with much larger temporal and
spatial resolution that CSI. |
|
| |
Figure 1: Displacement and
probability q-space images of healthy human
brain along with anatomical T1-IR image.
|
|
| |
Figure 2: Displacement and
probability q-space images of severe MS patient
along with FLAIR image showing the periventricular
lesions and NAWM. |
|
| |
2) High b value,
q-space analyzed diffusion imaging in Dementia
Alzheimer's disease (AD) and vascular dementia
(VaD) are the most common types of dementia (~50%
of dementia cases) and become a severe social
problem in developed countries. While the two
types of dementia leads to cognitive decline and
memory impairment, the nature of the two diseases
is different. In AD, the main pathologies are
observed in the cortical gray matter and are characterized
by the accumulation of neurofibrillary tangles
and senile plaques along with neuronal and synaptic
loss that produce cerebral atrophy. MRI can detect
the atrophy, which might be related to AD but
gives no specific diagnosis of the diseases. In
VaD, multiple ischemic lesions cause dementia
usually with a step-like course. In most cases
of VaD there is a gradual development of the disease
due to small-vessel involvement resulting in lacunar
infarctions in the basal ganglia and diffuse changes
in the white matter. These patients present diffuse
periventricular hyperintensities on T2-weighted
MRI known as leukoaraiosis. We have used high
b value q-space analyzed diffusion imaging to
follow white matter integrity in AD and VaD with
the aim to use this method to differentiate between
the two diseases as white matter is supposed to
be more damaged in VaD. Example for high b value
q-space analyzed images of an AD diseased brain,
VaD diseased brain and healthy control are shown
in the figure 3 below. In the AD case the high
b value q-space analyzed diffusion imaging showed
only little changes in the white matter. Most
of the white matter changes were observed in the
frontal lobe. Interestingly, the ability of the
q-space images to discriminate between gray and
white matter showed the shrinkage of the gray
matter due to atrophy while the white matter seems
to be only slightly damaged. In the VaD subjects,
however, a massive white matter abnormality, expressed
by a significant reduction in the apparent probability,
is observed in the q-space images, affecting almost
all areas of white matter. In contrast to AD the
gray matter in VaD seems to be intact. These results
are nicely seen in the blowup of the images shown
in figure 3 where the red color, which represents
mainly gray matter, seems to disappear in the
parietal lobe of the AD patient. However, in the
VaD subjects, the light blue areas, which represent
normal white matter, largely disappear suggesting
massive pathological changes in these white matter
areas. |
|
| |
Figure 3. q-Space probability
images of control, AD and VaD subjects. In the
second raw blowup images of the left parietal
lobe is depicted. Notice that in the control
subject, light blue represent normal white matter
and red represent normal gray matter. Areas
of gray matter seems to shrink in the AD patient,
while in the VaD subject it is the white matter
that seems to be massively damaged. |
|
| |
3. High b
value, q-space analyzed, diffusion imaging in
normal myelination
Brain development continues from birth to the
third decade of life. White matter maturation
is an important part in this process. Conventional
MR imaging can detect white matter maturation
following T1 and T2 relaxation times changes,
up to two years of age. Diffusion tensor imaging
shows major changes with age up to six years,
and only minor changes in adulthood. High b
value diffusion weighted imaging analyzed by
the q-space approach was shown to be very sensitive
to white matter maturation in rats. Maturation
of white matter includes both increase in the
diameter and myelination of axons, therefore
causes increase in restricted diffusion. In
this study we examined white matter maturation
from premature infants (36 weeks) up to 23-year-old
subjects. High b value q-space analyzed diffusion
weighted images demonstrated major changes in
white matter signal from infant to adults. Figure
4 shows probability images of six subjects at
different ages. At age of 4 month, almost no
contrast is seen between gray and white matter
suggesting that the contribution of the slow
diffusing component in the white matter is very
small. This might be related to low degree of
myelination at this age. As age increase and
myelin forms, the slow diffusing component becomes
more apparent and the probability for zero displacement
in white matter increases due to increase in
restricted diffusion. Interestingly, differences
can also be observed between age of 11 years
and an adult (22 years) especially in the sub-cortical
white matter. More than 17% changes in the probability
for zero displacement in observed between these
two age groups (9-11 years and 22-23 years).
This is in contrast to DTI where no changes
in diffusion anisotropy are observed over age
of 6 years. Indeed, it is known that some areas
in the brain reach full maturation only in the
second decade of life. It seems that q-space
image supports the order of maturation as observed
by conventional MRI but demonstrates changes
in white matter signal over a longer period
of life.
|
|
| |
Figure 4: q-Space probability
images for six subjects at different ages. Increase
in the probability for zero displacement observed
first in the corpus callosum and external capsula.
In older ages, increase in the probability for
zero displacement is observed mainly is sub-cortical
white matter. |
|
| |
4. DTI and fMRI
for evaluation of tissue functionally within and
surrounding brain lesions
In the last decade Functional MRI (fMRI) gradually
evolved to become a clinical tool for pre-surgical
evaluation of critical functionality in vicinity
of brain lesions. It aims to provide intervention
guidance and to minimize risk of functional deficit
following brain surgery. fMRI signals relate to
activated gray matter only. However the connectivity
of white matter is crucial for functionality of
a region and should not be damaged during surgery.
Diffusion tensor imaging (DTI) is a well-established
method for in-vivo mapping of the white matter
directionality and organization, but its clinical
application has not been fully explored, especially
in relation to functional reservoir in vicinity
of brain lesion. Our aim in this study was to
explore the added value of the combined methodologies
in brain surgery. In cases of cortical aberrations
(e.g. cortical dysplasia or hypotrophy) fMRI showed
distributed activation at regions that were suspected
as cortical aberration. DTI demonstrated organized
white matter approximating some of the activated
region but not at the dismorphed regions. By that,
DTI pointed to the functionally ineffective BOLD
responses in the candidate region for surgery
excision. Figure 1 shows a case of fronto-temporal
hypotrophy. In this case fMRI was indicated for
assessment of left hemisphere functionality. The
images were acquired from a 22y old right handed
female, suffered from intractable epilepsy and
mild mental retardation. fMRI paradigm consisted
of a language task of verb-generation in order
to determine laterality and distribution of frontal
and posterior canonical language regions. Activation
maps show active language representation in both
right and left hemisphere but more on right (see
arrows, Fig. 1B). DTI demonstrates well-organized
white matter in bilateral posterior regions, but
poorly organized fibers on left frontal region
(see arrows, Fig. 1C). This finding suggests that
in this case the fMRI activation shown in the
left Broca area might be functionally ineffective
suggesting that critical language is in the right
hemisphere (even though the patient was right
handed). In cases of space occupying lesions,
when fMRI activation were seen in vicinity of
lesion, DTI assisted in pointing to possible intact
fiber tracts leading to activation, therefore
possibly critical for functionality. Furthermore,
DTI could differentiate between cases in which
fibers were displaced by the tumor and could be
recovered by surgical intervention and cases of
tumor invading the functionally related fibers
and could not be recovered by surgery. In the
former, the outline of fiber tracks in relation
to fMRI activation has a significant added value
in terms of intra-operative navigation to avoid
the accidental cutting of functionally critical
fiber. Figure 2 shows a case of a space-occupying
lesion (left frontal meningioma). Functional mapping
was indicated for assessment of left hemisphere
functionality. The images were acquired from a
50y old right-handed female who was a candidate
for tumor excision. fMRI paradigm consisted of
a language task of verb-generation. Activation
maps show active language representation mainly
in the left hemisphere. The frontal (Broca) activation
was markedly displaced superior and posterior
to the expected location of the inferior frontal
gyrus (arrow, Fig. 2B). DTI map demonstrates well-organized
white matter displaced medially by the tumor but
is continuous to fibers approximating the language
related activated region (Fig. 2C). This finding
suggests that in this case the displaced BOLD
responses shown in left frontal region are most
probably functionally effective. |
|
| |
Figure 1: Fronto-temporal hypotrophy. (A) T1-weighted
image, (B) fMRI activation (white area), (C)
fractional anisotropy image.
|
|
| |
Figure 2: Left-frontal meningioma. (A) T1-weighted
image, (B) fMRI activation (white area), (C)
fractional anisotropy image |
|
| |
|
|
| |
|
|
| |
|
|
| |
|
|
| |
SORRY, THIS SITE IS STILL
UNDER CONSTRUCTION
|
|
