For my current thesis I had measurements of EEG activity from 24 mostly student subjects. The intention was to capture moments of fear or anxiousness which are induced only using audio effects. What turned out early that it's pretty hard to do. How can you arouse anxiousness to people that have different concepts of what's frightening and what's not? How can you point out the exact moment, when that feeling occurred? Are just audio inputs enough for someone to experience real danger?
I came up with the following solution: I coded a kind of dumb DJ set, that can play music one after the other automatically, meanwhile effects (short audio) can be played. When a music or an effect starts, a given message is sent on the given serial port that would mark the exact point on the EEG signal when (or after when) fear may occur. Volumes of music and effects can be adjusted and it makes a log file out of the played effects to remember order. Here's a picture of the GUI (implemented in C#):
So far so good. Now I can play frustrating base music meanwhile fear conditioning the subject having a 'condition' (as on the GUI) effect before playing the 'stimuli' effect. So the condition effect (conditional stimulus) would be a sound of a squeaking door, the stimuli effect (unconditional stimulus) would be an unpleasant scream for example.
To be able to differentiate signals at fearful moments from calm or neutral moments, I searched for a (proven to be) relaxing music beside a horror ambient music and some frightening effects.
Measuring equipment
- Emotiv EPOC neuroheadset and Emotiv TestBench software both provided by BUTE,
- Fear & anxiety measurement framework that I mentioned previously,
- EEGLAB Matlab toolbox for processing EEG signals,
- AKG earphone,
- my friend's scarf to cover eyes.
Course of a measurement
The measurements took place at a dark, quiet room as far as possible. The subject got decorated with the neuroheadset, earphone and scarf combo, then instructed not to move until I indicated it was over. The subjects were told that they were to experience a dream of a blind man, and they should find out where they were during listening.
After that, I started the calming music (~3 minutes) with some also calming natural quiet ambient effects in the background to guide the subject a little bit. The 'horror part' (~8 minutes) began sudden and harsh: loudly slammed iron door before a psychopathic scream. Then some effect pairs (conditional and unconditional stimulus) played on the fly, with approximately 40 seconds between these pairs.
Straight after measurement, the subjects were asked to rank on a scale of 1 to 10:
- the degree of relaxation at the beginning,
- the frustrating effect of the horror ambient music,
- the intensity of fear or anxiety effect by effect,
- and the overall degree of frustration in the second part.
These subjective rankings are used to filter out irrelevant EEG signals, I only analysed moments where effects were rated ≥ 7 that is happened to be 45 time points.
To be continued...
In the next post I'm going to present the analysis of the retrieved signals and observations of others on which I base my theory of fear measurement using EEG. Till that, here are some statistics I generated from my log files:
# Number of subjects: 24 (22 with log file) # Number of effects played: 272 # Male/female: 20/4 # Student/nonstudent: 18/6 # Avg age: 24.625 # Score distribution (0-10): ---------------------------- 0: 177 65.074% 1: 0 0.000% 2: 1 0.368% 3: 2 0.735% 4: 10 3.676% 5: 11 4.044% 6: 26 9.559% 7: 21 7.721% 8: 16 5.882% 9: 5 1.838% 10: 3 1.103% - it means that 21 + 16 + 5 + 3 = 45 points is to be analysed # Score distribution effect by effect: 0 1 2 3 4 5 6 7 8 9 10 ------------------------------------------------------------------------------------ scream_high: 1 0 0 0 0 0 0 0 0 0 0 cellar_door_slam: 11 0 1 0 3 1 2 2 1 0 1 dog_growling: 12 0 0 0 1 1 2 0 1 0 0 scream_distance: 6 0 0 0 1 0 0 0 0 0 0 female_finale_stimuli: 1 0 0 0 0 0 0 1 0 1 0 male_first_scream: 9 0 0 1 0 1 1 1 0 0 0 female_first_scream: 3 0 0 0 0 0 0 0 1 0 0 sqeaking_door: 15 0 0 0 2 0 5 0 0 0 0 laugh_closed_mouth_female: 14 0 0 1 0 0 0 0 0 0 0 footsteps_on_hollow_wooden_surface: 2 0 0 0 0 0 0 0 0 0 0 death_roman: 2 0 0 0 0 0 0 0 0 0 0 scream_psycho: 11 0 0 0 0 1 2 3 2 0 0 screams: 1 0 0 0 0 0 0 0 0 0 0 monster_growl: 11 0 0 0 1 1 2 0 0 1 0 scream_short_medium: 10 0 0 0 0 0 2 2 3 0 0 chainsaw: 1 0 0 0 0 0 0 0 0 0 0 chains_long: 15 0 0 0 0 2 2 0 2 0 0 finale_condition: 3 0 0 0 0 4 2 5 2 2 1 male_finale_stimuli: 7 0 0 0 0 0 3 2 3 0 1 zombie_animal_groan: 8 0 0 0 1 0 0 1 0 0 0 heartbeat: 9 0 0 0 0 0 0 1 0 0 0 slow_breathing: 9 0 0 0 0 0 1 1 1 1 0 long_laughter: 14 0 0 0 1 0 2 1 0 0 0 man_crying_heartbeat: 2 0 0 0 0 0 0 1 0 0 0 # Scoresum and times played of effects: scoresum times played rate ------------------------------------------------------------------------------ scream_high: 0 1 0.000 cellar_door_slam: 63 22 2.864 dog_growling: 29 17 1.706 scream_distance: 4 7 0.571 female_finale_stimuli: 16 3 5.333 male_first_scream: 21 13 1.615 female_first_scream: 8 4 2.000 sqeaking_door: 38 22 1.727 laugh_closed_mouth_female: 3 15 0.200 footsteps_on_hollow_wooden_surface: 0 2 0.000 death_roman: 0 2 0.000 scream_psycho: 54 19 2.842 screams: 0 1 0.000 monster_growl: 30 16 1.875 scream_short_medium: 50 17 2.941 chainsaw: 0 1 0.000 chains_long: 38 21 1.810 finale_condition: 111 19 5.842 male_finale_stimuli: 66 16 4.125 zombie_animal_groan: 11 10 1.100 heartbeat: 7 10 0.700 slow_breathing: 30 13 2.308 long_laughter: 23 18 1.278 man_crying_heartbeat: 7 3 2.333 - scoresum = sum of scores given - rate = scoresum / times played # Scoresum and times played of effect types: scoresum times played rate ----------------------------------------------------------------------------------- conditional stimulus: 349 160 2.181 unconditional stimulus: 260 112 2.321
No comments:
Post a Comment