Physiological sensors and EEG A short introduction to (neuro-)physiological measurements.

Physiological sensors and EEGA short introduction to (neuro-)physiological measurements

Outline

1. Physiological sensors and EEG

2. Classification of affective states

2.1 Background & experiment design

2.2 Features, Classification and Resumee

The sensorsElectroencephalogram (EEG)

Electroocculogram (EOG)

Electromyogram (EMG)

Electrocardiogram (ECG)

Galvanic skin response (GSR)

Blood volume pulse (BVP)

Respiration

Temperature

central nervoussystem

peripheral nervoussystem

Electro-----gram

..occulo..

..myo..

..cardio..

Blood volume pulse

infrared photoelectric sensor

detects changes in tissue blood volume

60 – 80 pulses a minute

> heart rate and inter-beat-interval can be derived

faster when exercised or aroused

Galvanic skin response

impedance of the skin is measured

eccrine glands at hands and food soles

increases linear with arousal

Respiration

circumference of chest is measured via strain gauge

normal breath rate: 12 - 16 breath per minute (~500ml)

faster and more shallow with higher arousal

Electroencephalogramnormal signal

theta band

alpha band

beta band

gamma band

0 – 1000 Hz

4 – 8 Hz

13 – 30 Hz

8 – 12 Hz

30 – 100 Hz

The 10 – 20 system

The lobes of the brain

Time- vs frequency domain

normal signal

theta band

alpha band

beta band

gamma band

0 – 1000 Hz

4 – 8 Hz

13 – 30 Hz

8 – 12 Hz

30 – 100 Hz

Event-related potentials (ERPs)

The raw signal with stimuli presentations

The averaging of the raw signal epochsleading to the ERP

ERPs from

affective pictures

Time-frequency analysis

evoked frequency analysis

induced frequency analysis

Tallon-Baudry & Bertrand, 1999

Recapitulation

EEG measures the electrical potentials from synchroneously active pyramidal cells in the cortex

high temporal resolution vs low spatial resolution

only weak signals from deeper structures & closed fields “blindness”

ERP analysis in time domain shows mainly low frequency components

analysis in frequency domain (evoked vs induced) also pics up high frequency components

Classification of

affective states

Background & experiment design

Motivation

Affective Computing aimes at the enrichment of human-computer interaction

Physiological and neurophysiological signals give access to (non-observable) affective and cognitive states

(Sander et al. 2005)

Emotion Models

2 dimensional model of emotion

valence

arou

sal

posneg

low

high

boredom

joy

fear

relaxation

An emotion, or an affective state, is a reaction to an internal (e.g. thought)or external event (e.g. visual or auditory stimuli), and a behavioural disposition.

frustration

Regulation

Affective State

Appraisal

Stimulus presentation

+/-

+/-

Hippocampus (HC)

dorsal Anterior Cingulate Gyrus (dACG)

dorsal prefrontal cortex (PFC)

Amygdala

Insula

ventral Anterior Cingulate Gyrus (vACG)

ventral prefrontal cortex (PFC)

Inside

Outside

HC

Insula

Amygdala

dACG

vACG

PFC

Neurobiology of Emotional PerceptionPhillips et al. 2003

Overview Issues aBCI

Elicitation of emotions

Ground truth construction

Sensor modalities

Modality fusion

Feature selection and reduction

Classification

Example Study

Elicitation of emotions

not easy to establish a natural context of emotion elicitation:

subject- vs event-elicited

lab setting vs. real world

expression vs feeling

open vs hidden recording

emotion vs other purpose (for subject)

Affective Stimuli

valence

arou

sal

posneg

low

high

boredom

joy

fear

relaxation

frustration

Ground truth construction

normed stimulus sets (e.g. IAPS, IADS)

versus

self-assessment after each trial

Sensor modalitiesphysiological modalities:

slow response to stimulus (seconds)

long inter-trial intervals (seconds)

few (but long presented) stimuli per subject

• neurophysiological modality:– fast response to stimulus (miliseconds)– short inter-trial intervals (miliseconds)– many (but short presented) stimuli per subject

Experiment: Design

Resting Stimulus SAM rating

3 – 4 s 6 s not limited

1. emotion induction with visual affective stimuli

2. self-assessment of (induced) affective state after each stimulus with Self-Assessment Maneken (SAM)

Classification of

affective states

Features, Classification and Resumee

Modality fusion

data-level fusion

feature-level fusion

decision-level fusion

Feature selection and reduction

Selection:manual selection (e.g. literature research)

automatical selection / “wrapper” method

Reduction:• mapping from high into low dimensional space

(e.g. PCA,ICA) > “filter” method

Feature extractionEEG: 6 features (from 6 frequency bands over 6 brain regions)

physiological Sensors: m

ean, variance, and

(minimum & maximum)

from GSR, blood pressure,

heart rate, respiration, and

temperature

Classification

• many possible classification approaches:decision trees, linear discriminant analysis, support vector machine, neural networks…

• optimal method depends on structure of data

• !keep training trials appart from test data, to avoid the contamination of the classifier.. !

Results

2 classes: high vs low arousal

3 classes: high, medium and low arousal

Classification• physiological studies:

– 4 class: up to 95%– 8 class: 81% (!)– generalization of classifiers over subjects and time

• neurophysiological studies:– 3 class: up to 67%– 2 class: up to 79% – ? generalization of classifiers over subjects and

time ?

Recapitulation

• natural context of emotion elicitation

• characteristics of sensors are important

• many different approaches for data fusion, feature selection/reduction and classification – no optimal method per se

• still a long way toward an affect classification in an natural and multimodal real-world setting.. but we are getting there!

Thanks