Transcript Imaging the Human Brain

Using Physics to Image the
Functioning Human Brain
Mark Mandelkern
University of California
Irvine
Roentgen 1895
• 1900: biomedicine relied almost entirely on
vision, hearing, touch, smell, taste.
• 2000: we observe Z (Xrays), acoustical
impedance (ultrasound), proton density and
spin-tissue interactions(NMR/MRI), electrical
activity in nerves and muscles (EEG/MEG/ECG),
tissue color (NIRS), physiological and chemical
processes (radioactive and magnetic tracer
methods), ….
• The Brain
–
–
–
–
Encased in a hard opaque shell
Isolated by the blood-brain barrier (BBB)
Nearly homogeneous in density
Doesn’t (directly) make noise, produce enzymes or
hormones
• Neurology/Psychiatry/Psychology (indirect)
– Around for >1000 years and still useful
– Investigation of the brain via its function
• Direct visualization
– Anatomy
– Supply of oxygen and nutrition (20% of body)
• Response to cognitive demand
– Electrical function
• Real-time neuronal activity
– Chemical function
• Multiple neurotransmitter systems
– Localization of sensorimotor, autonomic,
emotional, cognitive functions
EEG-MEG
• Current dipoles arise in the dendrite from
synaptic stimulation
– Na+, K+, Cl- currents pass through cell membrane
– EPSP (excitatory), IPSP (inhibitory)
• Each dendrite ~20 fAm for ~20 ms
• ~106 to yield an observable signal
• Skull surface evoked potentials ~ 0.1-10 mV
• Evoked magnetic field gradients ~ 0.01-1 pT/cm
• Time scale of evoked signals 20 to 200 ms
Fields of a current dipole
Q=Id
• EEG
• Surface evoked potential .1-10 mV
• Sensitive to conductivity of head
– Reduced by overlying conducting CSF and scalp
• Signal mainly from near-surface radial current dipoles
• Detected with low-noise, high impedance amplifiers
• MEG
• B from primary currents insensitive to head conductivity
– Ohmic currents produce additional B fields
• Signal mainly from transverse current dipoles
• Evoked fields 50-500 fT.
• Detected with DC SQUID gradiometers
• Electromagnetic Inverse problem (Helmholtz 1853 )
• Use E, B, anatomical data and simplifying assumptions to
infer the tissue current-dipole density
EEG cap
MEG gradiometers
SQUID array
MEG in shielded room
Computed Tomography
Xray Transmission CT
Positron Emission CT
X-ray CT Angiography
• With iodinated contrast media, real-time
observation of brain blood supply
• http://medical.toshiba.com/products/ct/Dyn
amicVolume/ClinicalCardiac01.aspx
Quantitative perfusion imaging with
diffusible tracers
• 133Xe - Single photon imaging
• H215O - PET (t1/2=2min)
• Arterial spin labeling (ASL); endogenous proton
spins - MRI
• Measuring the arterial concentration Cp(t) of
tracer and the tissue concentration VC(t) yields
the perfusion F(ml/g/min) and the distribution
volume V(ml/g).
A study of word and accent recognition with
neuropsychological testing, H215O PET and EEG
• Visual cues for specific words and accents followed 500ms later by
recorded spoken presentation in “dichotic” setting
• Presentation at 3 sec intervals for 12 2.5 min blocks separated by 10
minutes
• Results
– Recognition of words is faster but less accurate than accent recognition
(1000ms v. 1300ms) (73% v. 80%)
– Accent recognition engages a right hemispheric region homologous to
the left-hemispheric language region
– Perfusion and EEG findings agree with each other and with the
response-time result
• Conclusions
– There may be a brain region specialized to recognize like/unlike.
– It’s more important to accurately distinguish friend from foe than to get
words right.
Accent task – Word Task
Student t map for word/accent comparison
Magnetic Resonance Imaging
• NMR signal from proton magnetization
• Sensitive to free proton density and
– T1 (spin-lattice relaxation)
– T2 (intrinsic spin-spin dephasing)
– T2* (local B inhomogeneitydephasing)
– Other factors (spin motion, chemical effects..)
• Saturation recovery sequence
– Single excitation
– Spatial encoding of steady state signal
NMR saturation
recovery
NMR relaxation times
T1
T2
achieve
T1 weighted Spin-echo
T2 weighted
Spin-echo
T2* weighted
Gradient-echo
fMRI with BOLD
• Focal cerebral activity leads to local increases in blood
volume, perfusion and blood oxygenation
• These changes have a duration of 5-10s and affect a
brain volume larger than that electrically activated
• BOLD: Blood-oxygenation-level-dependent imaging
measures blood oxygenation by exploiting the
oxyhemoglobin/deoxyhemoglobin magnetic susceptibility
difference. Increased oxygenation causes a T2* increase
thus an increased signal
• Pauling measured c for both in 1936
• For RBC Dc~ -0.264 ppm (cgs)when oxygenated,
DB/B~ -0.45 ppm for venous blood when activation
Effect of blood oxygenation on
MR relaxation times
BOLD effect of blood
oxygenation in animal brain
100% O2
21% O2
Vascular response to activation
BOLD effect of visual
stimulation
Single-voxel response to hand
squeezing (BOLD-MRI)
Optical Functional Brain Imaging
• IR between 600 nm and
900 nm penetrates > 1 cm
into tissue
• NIRS at two wavelengths
distinguishes
oxyhemoglobin (strong
absorption at 830 nm) from
deoxyhemoglobin (strong
absorption at 758 nm)
• Problems: spatial
resolution and penetration
into the brain
fMRI Study of cigarette craving
• 42 heavy smokers abstinent for >20 min
• 9 fMRI blocks lasting 45 s each
– Neutral cue videos – 3 blocks
– Smoking cue videos - resisting craving
– Smoking cue videos – allowing craving
• Data-driven analysis-FSL
– Statistical parametric maps
BOLD-MRI effects of cigarette
craving I
Figure 2
Figure 1
Cigarette Cue
Crave >
Neutral Cue
4.5
Cigarette Cue
Resist >
Neutral Cue
2.3
2.3
Cigarette Cue
Crave <
Neutral Cue
8.0
2.3
5.4
Cigarette Cue
Resist <
Neutral Cue
5.0
2.3
BOLD-MRI effects of cigarette
craving 2
Figure 4
Positive
Correlations
5.2
Figure 3
Cigarette Cue Resist >
Cigarette Cue Crave
4.1
2.3
Negative
Correlations
4.9
2.3
Cigarette Cue Resist <
Cigarette Cue Crave
5.5
2.3
2.3
PET for probing the neurochemical
function of the brain
• Coincidence detection of 511 keV g rays
• Tomographic reconstruction of the distribution of
e+ emitter-labeled tracers
• Labels include 18F, 11C, 13N, 15O
• Tracers for glucose, amino acids, lipids,
perfusion agents, receptor ligands, ….
• Example: 18F-2FA, a high affinity, selective,
nicotinic acetylcholinergic antagonist
PET-18F-2FA study of nicotinic
acetylcholine receptor binding following
smoking
H
N
H
O
H
N
F
• 14 heavy smokers
• 18F-2FA bolus plus infusion for 4 hours
• Subjects smoked 0, 1 puff, 3 puffs, 1
cigarette, 2-3 cigarettes
• Infusion and scanning for 4 more hours
• Measurement of activity in thalamus, brain
stem, cerebellum
18
MRI
baseline
kBq/ml
9
after
smoking
0
0.0 cig
0.1 cig
0.3 cig
1.0cig
3.0cig
NonDisplaceable
Time activity curves without
smoking
Radiactivity, kBq / ml
10
8
6
4
Th
Bs
Cb
2
0
0
60
120
180
240
300
Time, min
360
420
480
Time activity curves after
smoking
125%
125%
100%
2
75%
3
4
5
50%
25%
1
Radioactivity
thalamus
Radioactivity
1
100%
2
75%
3
4
5
50%
brainstem
25%
A
0%
B
0%
0
120
240
Time, min
360
0
120
240
Time, min
360
125%
cerebellum
Radioactivity
1
100%
2
75%
3
4
5
50%
25%
C
0%
Radioactivity, kBq / ml
8
Th
Bs
Cb
Frcx
Br tot
Cc
6
4
2
D
0
0
120
240
Time, min
360
0
60
120 180 240 300
Time, min
smoking
to satiety,
2.8 cig
Conclusions
• The brain is extraordinarily complicated.
• We have a broad assortment of
complementary tools for exploring its
function.
• We will continue to be confused and
surprised by what we find.
• This activity will keep us busy for a long
time.