A linear relationship between

1
A linear relationship between synaptic activity
and hemodynamic responses in brief stimulation
revealed with simultaneous recordings of local
field potentials and optical intrinsic signals

Masahito Nemoto
Laboratory of
NeuroImaging
Department of
Neurology
UCLA School of Medicine

Los Angeles, California, USA
2
Experimental Plan

• Specific Aim
• How metabolic and hemodynamic response
  signals
• evolve in time and space ?
• What type of response signals has higher
  fidelity to
• measured neuronal signals ?
  
• Hypothesis
• Optical response signals evoked by several
  pulses
• are built of unit responses in linear
  fashion.

3
Methods

• Male Sprague-Dawley Rats (n 9)
• Thinned Skull
• Enflurane Anesthesia
• Hind Paw Electrical Stimulation
• CCD Imaging (570 nm, 610 nm filter)
• Microelectrode Insertion (0.2-0.4 M?, 500 ?m
  depth)
• Simultaneous Recordings of
• Local Field Potentials and Optical
  Intrinsic Signals

4
Electrophysiology
Which components of neuronal activity contribute
to the hemodynamic response ?

• Functional blood flow and BOLD responses are
  better correlated with local field potentials
  than single- and multi-unit activity
• (Lauritzen, Mathiesen, Nielsen et al., 1998,
  2000, 2001 )
• (Logothetis et al., 2001)

Summed Evoked Field Potential ?FP
5
Optical Imaging of Intrinsic Signals
570 nm Isosbestic Point of Oxy- and DeoxyHb 610
nm DeoxyHb Dominant Wavelength
6
Stimulation Protocols

• Protocol 1 (5 pulses, 200 ms ISI)
• variable stimulus intensity
• 0.4 0.8 1.6 mA
• Protocol 2 (0.8 mA, 200 ms ISI)
• variable stimulus number
• 1 3 6 pulses
• Protocol 3 (0.8 mA, 2 pulses)
• variable interstimulus interval (ISI)
• 200 150 100 66 ms

Microscopic scale ? 1 sec
7
Linear Systems

• Homogeneity Property
• Additivity Property
• Superposition Property

f (a x) a f (x)
8

• Long stimulus duration

The Linear Transform Model Temporal
integration of the responses to short stimuli
based on time-invariant systems.

• Short stimulus duration

Without temporal integration, we simply
compared ?FP with the peak optical responses.
Impulse response
9
RESULTS
Yosemite Valley
10
Maximum Response 4-size ROIs
C1
11

ROI center-to-center distance between time
epochs and wavelengths
ROI in early phase at 570 and 610 nm
570 nm ROI
610 nm ROI
Center-to-center distance (n 32) 570 nm
early and peak 158 ? 190 ?m
610 nm early and late peak 930 ? 352 ?m
570 nm and 610 nm early 196 ? 147 ?m
12
ROI determination and size dependence of
optical signal at 570 and 610 nm
Standard Stimulation
Stimulus Intensity 0.8 mA Stimulus
Number 5 pulses Interstimulus
Interval 200 ms
13
Protocol 1 Stimulus intensity dependence of
optical signal at 570 and 610 nm
14
Protocol 2 Stimulus number dependence of
optical signal at 570 and 610 nm
15
Protocol 3 Interstimulus interval dependence of
optical signal at 570 and 610 nm
16
Spatiotemporal Profiles of Optical Signals at 570
and 610 nm
17
Protocol 1 Variable Stimulus Intensity
18
Protocol 2 Variable Stimulus Number
19
Protocol 3 Variable Interstimulus Interval
20
Correlation analyses between ?FP and optical
response magnitude
in average data
21
Scatterplots of normalized ?FP versus optical
responses A representative case in Protocol 1
22
Protocol 1 (variable stimulus intensity) Normalize
d ?FP vs. Optical responses
9 rat data on a trial-by-trial basis
23
Protocol 1 (variable stimulus number) Normalized
?FP vs. Optical responses
9 rat data on a trial-by-trial basis
24
Protocol 3 (variable ISI) Normalized ?FP vs.
Optical responses
9 rat data on a trial-by-trial basis
25
Total Data (n1305 trials) Normalized ?FP
vs. Optical responses
9 rat data on a trial-by-trial basis
26
Fidelity of the optical responses to minute ?FP
fluctuations within the same stimulus
Distribution of determination coefficients (r2)
in 27 trial sets of Protocol 1 across 9 rats
Time series (A representative case)
27
610 nm early negative vs. late positive
signals
28
Non-Linearities
We observed several types of systematic
departures from linearity associated with the
analysis methods and stimulation protocols.
29
Different slopes between protocol-1 and -2
30
CONCLUSION
Grand Canyon
31
CONCLUSION
570 and 610 nm spatiotemporal evolution patterns
and linear relationships with SFP indicate
CBV-related hemodynamic responses have higher
fidelity to integrated synaptic activity than
oxygenation-derived signals, and support the
validity of linear systems analysis of these
responses to evaluate neuronal activity.
32
Acknowledgements
Sameer Sheth Michael Guiou Nader
Pouratian James Chen Nathan Hageman
Melissa Walker
UCLA Neuro Imaging Lab. Dr. Arthur W Toga
33
(No Transcript)
34
(No Transcript)
35
Imaging Analysis

• ratio (t) I (t) / I base
• ratio (t) ratio (t) ratio (-0.25)
• I base I (-1.0) I (-0.75)
• I (-0.5) I (-0.25)
  / 4

• The magnitude of optical intrinsic signals was
  calculated as
• the fractional change in reflected light
  intensity relative to
• the pre-stimulus baseline.
• Pixel-by-pixel division and subtraction
• ? ratio and ? absorbance are almost
  proportional the ratio
• between them varies by only 2.6 in the
  response range (0-5).

36
Correlation analyses between ?FP and peak areal
extent
37
Oscillations of 577 and 605 nm reflected
light induced by hypotension
577 nm
605 nm
Oscillation Frequencies 6 cycles / min
38
Oscillations of reflected light at 605 nm induced
by hypocapnia
0 sec
1.0
2.0
3.0
4.0
A
B
39
Oscillations of Cortical Bands 
40
Specific Aim I Hypothesis 1
W
HL
41
Specific Aim I Hypothesis 1
610 nm early
610 nm late
570 nm
SEP
42
Specific Aim I Hypothesis 1
610 nm early
610 nm late
570 nm
2 Hz
5 Hz
10 Hz
15 Hz
3.1
14
11
20 Hz
W
5.8
51
46
HL
43
SUMMARY

• 570 nm optical intrinsic signals centered on
  activated pial arterioles. Their spatial geometry
  was similar at different time points and across
  graded stimulation. The spatial pattern of 610 nm
  signals metamorphosed over time from capillary
  bed to large draining veins.

2. Correlation analyses revealed significant
linear relationships between summed evoked FP
(?FP) and both 570 and 610 nm signals. On a
trial-by-trial basis, the determination
coefficients of 570 nm signals were considerably
higher than those of 610 nm signals, and 570 nm
signals even correlated with minute ?FP
fluctuations within trials of the same stimulus.
3. We also recognized systematic nonlinear
components associated with hemodynamic response
regulation, analysis procedures, and stimulation
protocols.