Brain connectivity
Task 2
Do we need to study connectivity to understand brain function
Why are there so many connectivity studies lately
How does neural tracing work
How can we study white matter
Why can't we use it in humans
How can we see what is connected to what
How can we study human wiring with MRI
What is resting state functional connectivity
What is DTI
If 2 structures correlate, does that mean they are connected?
invasive and destructive
Based on signals from 1H(proton) nuclei
limitations
spatial resolution
contrast
10 mumeter
data size
too big or small?
Water
if these molecules are the same then regions indestinguishable
Contrasts
dependent on
Proton density (PD)
T1 and T2
Diffusion coefficient (D)
decay times after excitation
motion of water molecules
formula for singal
S=PD(1-e2TR=T1 )e2TE/T2e-bD
b
diffusion weigning factor
can be changed
important for DTI
D
diffusion term
motion of water molecules
3 facts
water molecules move
DTI uses this motion to infer neuroanatomy
DTI is mostly about static anatomy
diffusion
isotropic
circular ink stain
anisotropic
when higher fiberdencity is oriented in one direction
estimate axonal organization
B0 field
strong magnetic field along bore
can be linearly altered by gradient pulse
Gradients
X Y Z
changing strength and timing of gradient pulses
click to edit
detects water along gradient axis
Combine X, Y, Z axes
fiber orientations mostly oblique to axes
FA
Fractional anisotrophy
0(isotropic) to 1(anisotropic)
limitations
prone to artifacts caused by physiological motion
limited spatial resolution
Advantages
2 types of new contrasts
diffusion anisotropy
fiber orientation
can display the complexity of white matter
cannot differentiate directionality of axons
diffusion can be blocked by obstacles
structures are assumed homologues within pixel
Application
check myelinization
T2 contrast
reveal locations of white matter bundles
click to edit
studying the level of coactivation between functional time-series of anatomically separated brain region during rest
Patterns usually reflect functional correlation
How to process this data
model dependent methods
seed method
model free methods
correlate resing state of 1 region against the rest of the brain
functional connectivity map
must be defined a priori
simple analyses and straight foreward result
difficult to examine whole brain scale
Examines whole brain
Principal component analysis
independent component analysis
searches mixture of underlying sources that can explain pattern
add clustering strategies
all methods have high overlap
Resting state networks
anatomically separated but functionally linked regions of functional connectivity during rest
8 subnetworks
motor
visual
lateralized superior parietal /frontal
default mode
bilateral temporal/insular and ACC
Analyses
small world principle
click to edit
graph theory
click to edit
Disorders
Alzheimers
decreased default mode
decreased resting activity in PCC and hippocampus
schizophrenia
aberrant default mode connectivity
decreased connectivity in medial frontal cortex and precuneus
altered integrity of the cingulum
MS
click to edit
ALS
inject a tracer to see where it goes
retrograde and antrograde tracing
cons
brain has to be taken out
Pros
good technique
Signal loss between pulses = diffusion
many different gradients
for each direction
ellipsoid
around the points were it diffused to
the pointier the more diffusion
limitations
influenced by things like breathing