Imaging the Human Mind

1
Imaging the Human Mind

• Basics of Magnetic Resonance Imaging

2
Moving electric charge (i.e. electric current)
produces magnetic field

• Current sink
• ?
• B
• ?
• Current source
• (Right hand rule)
• Electromagnetic Radiation has many names
• at different frequencies (energy levels) it is
  called radio, microwave, light, x-rays, gamma
  rays, etc.

3
And vice versaMoving or changing magnetic
field gt electric force electric force (i.e.
voltage difference) gt electric current
4
(No Transcript)
5
An MRI scanner magnetic field is usually created
by producing an electrical current in a tube of
superconducting material
e-
B
6
(No Transcript)
7
1 Tesla 10,000 Gauss or 20,000 times the
earth's magnetic field.Most clinical scanners
are 1.5 Teslamost fMRI research scanners are
1.5 or 3.0 Teslaanimal scanners and a couple of
human scanners 7T, 10T or even moreThe stronger
the magnetic field, the stronger the MRI signal.
More on why later
8
Protons have positive electric charge and "spin"
which results in their behaving like little
dipole magnets.
Because remember a moving electric charge (i.e.
electric current) produces magnetic field
9
When in an external magnetic field, their lowest
(preferred) energy state is to line up with the
field.
B
But only slightly more than 1/2 of the protons
will actually line up, the others will be in the
opposite orientation. The stronger the field, the
more will line up.
10
(No Transcript)
11
During an MRI scan, a radiofrequency pulse is
applied that knocks some protons out of alignment
with the magnetic field
B
Then the protons have to start lining up all over
again.
12
While out of alignment, the proton emits
radiofrequency EM radiation that is detected by
the scanner
Spinning proton in a magnetic field (or a
spinning top in a gravitational field) will
"precess"
longitudinal component of the dipole
Transverse component of the dipole
13
Spinning proton in a magnetic field (or a
spinning top in a gravitational field) will
"precessat a speed determined by the strength of
the magnetic field f ?Bo
14
The transverse component of the proton dipole is
the source of the MRI signal.
longitudinal component of the dipole
Force (Bo)
Transverse component of the dipole
15
(No Transcript)
16
(No Transcript)
17
(No Transcript)
18
(No Transcript)
19
MRI is all about phase and frequency
f ?Bo

• where f is the frequency of precession (Larmor
  frequency),
• is the Larmor constant (aka gyromagnetic ratio),
  and
• Bo is the strength of the magnetic field

Changes in the magnetic field can be externally
applied in order to alter the phase and/or
frequency at different locations in the brain (or
body) in order to figure out where the signals
are coming from.
20
Making a 3D Image out of radio signals
21
B field gradient determines the frequency at
which protons will resonate. Apply pulse with a
restricted frequency range, tip protons only in
one slice.
22
Making a 3D Image out of radio signals
23
Phase Encode (applied before readout)
Frequency Encode (applied at time of readout)
24
MRI Contrast is Flexible!
WaterConcentration
Dephasing
Magnetization Recovery
25
Basic types of MR ImagesDifferent types of
Contrast between tissue types

• Proton Density
• T1-weighted
• T2-weighted
• T2-weighted

26
MRI is all about phase and frequency
f ?Bo

• Even without applying gradients,
• the magnetic field is not the same strength
  everywhere.
• It varies on both a macroscopic and microscopic
  scale.

27
Transverse component (MRI signal)gets smaller for
2 reasons, relaxation (T1) and dephasing (T2),
which depend on different properties of the tissue
28
T1 f(Longitudinal relaxation time)
B
The process of lining up with the field is called
longitudinal relaxation. The time it takes for
the this portion of the protons get lined up with
the magnetic field is a function of "T1" - the
smaller (shorter) T1 is, the shorter the time to
line up (become "magnetized"). T1 is different
in different types of tissues (water long T1,
fat short T1).
29
Transverse component (MRI signal)gets smaller for
2 reasons, relaxation (T1) and dephasing (T2),
which depend on different properties of the tissue
30
T2 and T2 f(time to get out of phase)
There are lots of protons each giving off their
own signal
31
(No Transcript)
32
T2 and T2 f(time to get out of phase)
If they are all at the same point in their
precession cycle, (I.e. in phase) then their
signals will add together.
33
T2 and T2 f(time to get out of phase)
If they are all at different points in their
precession cycle, (I.e. out of phase) then
their signals will cancel each other.
34
Source Mark Cohens web slides
http//porkpie.loni.ucla.edu/BMD_HTML/SharedCode/M
iscShared.html
35
But why would they get out of phase?

• Because the magnetic field is not the same
  strength everywhere.
• It varies on both a macroscopic and microscopic
  scale.

f ?Bo
T2 depends on the microscopic variations (due
to chemical bonds, molecular structure, etc) T2
depends on the macroscopic variations (due to
tissue density, hemoglobin state, etc)
36
T1 vs. T2
Weighting T1 or T2 or T2 means you have
maximized the contrast between tissues according
to that property and minimized contrast according
to other properties
MRI signal magnitude is a function of both the T1
and T2 properties of the tissue at that location
Source Mark Cohens web slides
37
(No Transcript)
38
TE Time to echo (read signal) TR Time to
repeat
39
T1 and TR

• T1 recovery of longitudinal (B0) magnetization
• T1 contrast used in anatomical images
• 500-1000 msec (longer with bigger magnetic
  field)
• TR (repetition time) time between excitation
  pulses

Source Mark Cohens web slides
40
T2 and TE
T2 decay of transverse magnetization TE (time
to echo) time to wait to read signal after
excitation pulse
Source Mark Cohens web slides
41
T1
T2
42
T1 vs. T2
Source Mark Cohens web slides
43
T2

44
Example of T2 weighted image
EPI vs Spiral Activation Slide from Gary Glover
OHBM meeting 2000
5 mm/skip 0 mm
5 mm/skip 2.5 mm
5 mm/skip 5 mm
EPI
0.2 r 0.4
Spiral
0.4 r 0.7
45
How do you get multiple slices?
After reading signal for 1st slice but before the
next excitation of that same slice, excite
another slice and read it, and so on for as many
TEs as you can fit into the length of the TR.
(TR/TE slices) Need to interleave slices or
leave gap to avoid interference between slices.