Dick J' Bierman

1
(No Transcript)
2
.
Anomalous anticipatory brain activation preceding
exposure of emotional and neutral pictures An
fMRI Study

• Dick J. Bierman Steven H. Scholte
• Universities of Amsterdam Utrecht (NL)

PA Convention, August 5-8, Paris, France
3
PRECognition or Sentiment
4
What is Presentiment?
Consider the sep. 11 WTC attack. Was there
anybody in the general public having some uneasy
feeling before the attack?
5
There are a lot of stories.
6
Was this just coincidence?
Ed Cox (1960) Compared train ridership on days
with and without a train accident. He found that
on accident days there were significant less
people on trains. (Note Utts redid the analyses
and found a smaller effect after correction for
holidays)
7
What went on before? Presentiment experiment
Presentiment is reported as the apparent
psycho-physiological effect of a future emotional
cause.
8
Procedure

• Subject sits relaxed with electrodes attached
• Gets about 40 exposures of randomized neutral and
  calm stimuli
• Skin conductance is averaged separately for
  emotional and neutral stimuli
• Before, during and after stimulus
• Baseline fixed at -7.5 seconds

9
Procedure Presentiment trial
Skin Conductance
Time
10
For example the following sequence could arise
c
c
c
c
c
e
e
e
We analyze this by averaging over the condition
11
Results First Subject (Radin-1)
Before
During
After
12
Results all Subjects (Radin-1)
Emotional
Calm
13
Replications
14
Replications

• Amsterdam University
• Edinburgh University
• Lab. For Fundamental Research, Palo Alto
• Cornell University (in progress)
• Budapest
• POST HOC in already published data
• Antonio Damasio (Iowa university)
• Alfons Hamm (Uni.Greifswald)

15
This should be everywhere!
Main stream emotion research
Measure baseline of dependent var Present an
emotional / calm event Measure response of
dependent var
baseline
event
response
time
16
Mainstream Research questions
R f ( emotional level of event) Find f R
Response - BaseLine Implicit assumption BaseLine
is independent of future emotional level
17
What did we do?
Search for main stream databases that Measure a
BaseLine for at least 4 seconds. Use strong
emotional events For which Data can be obtained
18
What did we find?
Hamm group on the speed with which fear arises
in animal phobic patients Damasio group on
implicit emotional learning during Gambling task
19
Procedure Animal fear study
Skin Conductance
time
20
Results re-analysis Hamms data
stimulus
21
Damasio gambling procedure

• Participant gets initially 2000
• Draws cards from one of 4 different decks
• Card is either winning or losing

Preparation
Skin Conductance
time
22
Damasio dataset analyses

• Damasio-analysis P f (type of deck) !!!
  Advantageous vs non-advantageous
• Our Re-analysis P f(type of Card)
  Winning vs Losing

23
Winning vs. Losing Cards

• Quote from Bechara, Damasio et al, Science,
  275, 28 February 97, 1293-1295
• .. the players have no way of predicting when
  a penalty will arise

24
Results re-analysis Gambling experiment

• t 1.634 df117 p 0.053
• Presentiment effect 20!!!!

25
So What Next?

• Can we differentiate between different emotions?
• Can we locate the source of this phenomenon?

26
Brain scanner, Amsterdam Summer 2001
27
Event Related Design

• 10 subjects
• 48 pictures (events)
• 32 neutral, 8 erotic, 8 violent
• 4.2 seconds exposure time
• 16.8 seconds for one trial ( 8 volumescans)
• 1 vol 22 slices, slice resolution 64 64

28
Analysis procedure (1)

• Dependent var BOLD signal per voxel (Blood
  Oxygenation Level Dependent)
• Pre-processing
• Slice time correction
• Linear trend removal
• 8 mm spatial smooting
• No further smoothing in time domain

29
Analysis Procedure

• SPATIAL
• Fit data to a General Linear Model
• Predictors neutral, erotic, and violent stim.
• Result Regions of Interest with significant fit.
• TEMPORAL (restricted to ROI)
• Event related averaging of BOLD per condition
• t-test of average BOLD signals in ROI

30
(No Transcript)
31
Screen Dump of analysis software Output Input
Single individuals brain signals
32
Erotic Presentiment effect at time -4
seconds. Td 2.89 df 39 p lt 0.01
33
(No Transcript)
34
(No Transcript)
35
Amygdala
Negative presentiment in neutrals???
36
Pooling of data

• Mapping in Talairachspace
• Loss of signal due to individual differences
• Soseparate male and female

37
all female
38
all male
39
Considered Normal explanations

• Artefacts Introduced by software
• Hemodynamic curve
• Smoothing
• Chance fluke or over analysis
• Incorrect Randomization
• Subject Strategies

40
Subject Strategies
A subject strategy reflects itself in the
anticipation being a systematic function of the
history of stimuli so far presented.
41
Subject anticipatory Strategies

• Subject strategy Ai f(P1P i-1)
• Based upon Gamblers Fallacy
• In spite of the fallacy produces normal effect
• Goes to 0 if N goes to infinity
• Simulation using ACTUAL random sequence
• A increases after each neutral (sigmoid function)
• A is reset after emotional stimulus

42
Results SS simulation
Over-all Subjects the explained effectsize is 2
after N48 trials
43
Reasons to reject SS explanation

• Explained effect is too small
• Does not explain spatial qualitative differences
  between emotions and gender
• NO neutral anticipation observed!!!!!

44
Conclusion

• This study should be seen as exploratory
• But justifies further research (as usual)
• And theoretical speculation
• Could Consciousness be a Feynman-Wheelers
  multiparticle coherent absorber and therefore
  allow for advances waves?
• Replication study will focus on the coherence of
  brain states