« Brain-Computer Interfacing

Klaus-Robert Müller

Brain Computer Interfacing (BCI) aims at making use of brain signals for e.g. the control of objects, spelling, gaming and so on. This tutorial will first provide a brief overview of the current BCI research activities and provide details in recent developments on both invasive and non-invasive BCI systems. In a second part – taking a physiologist point of view – the necessary neurological/neurophysical background is provided and medical applications are discussed. The third part – now from a machine learning and signal processing perspective – shows the wealth, the complexity and the difficulties of the data available, a truely enormous challenge. In real-time a multi-variate very noise contaminated data stream is to be processed and classified. Main emphasis of this part of the tutorial is placed on feature extraction/selection, dealing with nonstationarity and preprocessing which includes among other techniques CSP. Finally, I report in more detail about the Berlin Brain Computer (BBCI) Interface that is based on EEG signals and take the audience all the way from the measured signal, the preprocessing and filtering, the classification to the respective application. BCI communication is discussed in a clinical setting and for gaming.

Scroll with j/k | | | Size

Slide: Brain Computer Interfacing

Klaus-Robert Mller, Carmen Vidaurre, Matthias Treder, Siamac Fazli, Jan Mehnert, Stefan Haufe, Frank Meinecke, Felix Biessmann, Michael Tangermann, Gabriel Curio, Benjamin Blankertz et al.

Brain Computer Interfacing

Klaus-Robert Mller, Carmen Vidaurre, Matthias Treder, Siamac Fazli, Jan Mehnert, Stefan Haufe, Frank Meinecke, Felix Biessmann, Michael Tangermann, Gabriel Curio, Benjamin Blankertz et al.

1

Slide: BCI MLSSP 2012 Topics
Part I - Physiology, Signals and Challenges - Event-Related Desynchronization and BCI

Part II - Nonstationarity SSA et al. - Multimodal data

Part III - Event Related Potentials and BCI

- Applications

BCI MLSSP 2012 Topics
Part I - Physiology, Signals and Challenges - Event-Related Desynchronization and BCI

Part II - Nonstationarity SSA et al. - Multimodal data

Part III - Event Related Potentials and BCI

- Applications

2

Slide: Some BCI Groups (not an exhaustive list!) from !2003!
Schwarz, Pittsburg: Invasive Chapin, Rochester: Invasive  Pfurtscheller, Graz: ERD, Patients  Bayliss, Rochester: P300 & VR

Nicolelis, Duke: Invasive
Kennedy, Atlanta: Invasive Levine, Michigan: Invasive

 Penny, Roberts, Sykacek, Oxford: Bayes & BCI
 Birch and Mason, UBC BCI  Moore, Georgia BCI

Wolpaw, Albany: BCI 2000, 2D, Patients
Donchin, Beckman: P300: Spelling Anderson, UC, CSU: NN for BCI, invasive Sadja and Parra, NY: SP, Rapid Visual Stimulation Birbaumer, Kbler T: SCPs, TTD, Patients

 Allison, UCSD
 Millan, EPFL: brain states control robot  Donoghue, Brown U, invasive patient study  Cuntai Guan, Singapore: P300  Gao, Beijing: P300  BBCI: Let the machines learn

!Note that this is historical slide and MOST groups are missing!

Some BCI Groups (not an exhaustive list!) from !2003!
Schwarz, Pittsburg: Invasive Chapin, Rochester: Invasive Pfurtscheller, Graz: ERD, Patients Bayliss, Rochester: P300 & VR

Nicolelis, Duke: Invasive
Kennedy, Atlanta: Invasive Levine, Michigan: Invasive

Penny, Roberts, Sykacek, Oxford: Bayes & BCI
Birch and Mason, UBC BCI Moore, Georgia BCI

Wolpaw, Albany: BCI 2000, 2D, Patients
Donchin, Beckman: P300: Spelling Anderson, UC, CSU: NN for BCI, invasive Sadja and Parra, NY: SP, Rapid Visual Stimulation Birbaumer, Kbler T: SCPs, TTD, Patients

Allison, UCSD
Millan, EPFL: brain states control robot Donoghue, Brown U, invasive patient study Cuntai Guan, Singapore: P300 Gao, Beijing: P300 BBCI: Let the machines learn

!Note that this is historical slide and MOST groups are missing!

3

Slide: Increasing Interest by Scientists

Courtesy of Dr. Jon Wolpaw, Wadsworth Center

Increasing Interest by Scientists

Courtesy of Dr. Jon Wolpaw, Wadsworth Center

4

Slide: The origins of EEG and MEG (short recap.)
EPSPs and IPSPs

[From Vigario]

The origins of EEG and MEG (short recap.)
EPSPs and IPSPs

[From Vigario]

5

Slide: From single units to patch of dipoles

[From Vigario]

From single units to patch of dipoles

[From Vigario]

6

Slide: From single units to patch of dipoles (cont.)

[From Vigario]

From single units to patch of dipoles (cont.)

[From Vigario]

7

Slide: A glance at the cerebrum Motor cortex

[From Vigario]

A glance at the cerebrum Motor cortex

[From Vigario]

8

Slide: From dipole patches to EEG

[From Vigario]

From dipole patches to EEG

[From Vigario]

9

Slide: Invasive vs noninvasive Brain Computer Interfacing

[From Schalk]

Invasive vs noninvasive Brain Computer Interfacing

[From Schalk]

10

Slide: Invasive BCI at its best

[From Schwartz]

Invasive BCI at its best

[From Schwartz]

11

Slide: ECOG

 presurgical localization of area causing epilepsy

 excellent possibilty to learn about brain for human subject

[From Schalk]

ECOG

presurgical localization of area causing epilepsy

excellent possibilty to learn about brain for human subject

[From Schalk]

12

Slide: Invasive vs noninvasive Brain Computer Interfacing

[From Birbaumer et al., Nicolelis et al]

Invasive vs noninvasive Brain Computer Interfacing

[From Birbaumer et al., Nicolelis et al]

13

Slide: Noninvasive Brain-Computer Interface

DECODING

Noninvasive Brain-Computer Interface

DECODING

14

Slide: Brain Pong with BBCI

Brain Pong with BBCI

15

Slide: Noninvasive BCI: clinical applications
Brain- Computer Interface
Signal Processing
EEG Acquisition Application Interface

FES Device Grasp-Pattern 3 channel Stimulation

[From Birbaumer et al.]

[From Pfurtscheller et al.]

BBCI: Leitmotiv: let the machines learn

Noninvasive BCI: clinical applications
Brain- Computer Interface
Signal Processing
EEG Acquisition Application Interface

FES Device Grasp-Pattern 3 channel Stimulation

[From Birbaumer et al.]

[From Pfurtscheller et al.]

BBCI: Leitmotiv: let the machines learn

16

Slide: The cerebral cocktail party problem

 use ICA/NGCA projections for artifact and noise removal  feature extraction and selection

[cf. Ziehe et al. 2000, Blanchard et al. 2006]

The cerebral cocktail party problem

use ICA/NGCA projections for artifact and noise removal feature extraction and selection

[cf. Ziehe et al. 2000, Blanchard et al. 2006]

17

Slide: Towards imaginations: Modulation of Brain Rhythms

Single channel

IMAGINATION of left arm

Towards imaginations: Modulation of Brain Rhythms

Single channel

IMAGINATION of left arm

18

Slide: Variance I: Single-trial vs. Averaging

Single channel

Variance I: Single-trial vs. Averaging

Single channel

19

Slide: Variance II: Session to Session Variability

maps

Variance II: Session to Session Variability

maps

20

Slide: Variance III: inter subject variability [l vs r]

Variance III: inter subject variability [l vs r]

21

Slide: BCI with machine learning: training

BCI with machine learning: training

22

Slide: BBCI paradigms
Leitmotiv: let the machines learn

- healthy subjects untrained for BCI

A: training 20min: right/left hand imagined movements  infer the respective brain acivities (ML & SP)

B: online feedback session

BBCI paradigms
Leitmotiv: let the machines learn

- healthy subjects untrained for BCI

A: training 20min: right/left hand imagined movements infer the respective brain acivities (ML & SP)

B: online feedback session

23

Slide: BBCI paradigms
Leitmotiv: let the machines learn
- healthy subjects (BCI untrained) perform "imaginary movements (ERD/ERS)
- instruction: imagine

- squezzing a ball, - kicking a ball, - feel touch

BBCI paradigms
Leitmotiv: let the machines learn
- healthy subjects (BCI untrained) perform "imaginary movements (ERD/ERS)
- instruction: imagine

- squezzing a ball, - kicking a ball, - feel touch

24

Slide: Playing with BCI: training session (20 min)

Playing with BCI: training session (20 min)

25

Slide: Machine learning approach to BCI: infer prototypical pattern

Inference by CSP Algorithm

Machine learning approach to BCI: infer prototypical pattern

Inference by CSP Algorithm

26

Slide: Average topology of idle SMR

Average topology of idle SMR

27

Slide: Spatial Smearing

Spatial Smearing

28

Slide: The need for spatial filtering

The need for spatial filtering

29

Slide: Analysis of motor imagery conditions: spectra

Analysis of motor imagery conditions: spectra

30

Slide: ERD curves of motor imagery

ERD curves of motor imagery

31

Slide: Common Spatial Pattern Analysis

Common Spatial Pattern Analysis

32

Slide: Common Spatial Patterns for 2 classes

[cf. Blankertz et al. 2008, Lemm et al. 2005, Dornhege et al. 2006, Tomioka & Mller 2010]

Common Spatial Patterns for 2 classes

[cf. Blankertz et al. 2008, Lemm et al. 2005, Dornhege et al. 2006, Tomioka & Mller 2010]

33

Slide: CSP at work

CSP at work

34

Slide: Distribution of EEG features

Distribution of EEG features

35

Slide: BBCI Set-up

Artifact removal

[cf. Mller et al. 2001, 2007, 2008, Dornhege et al. 2003, 2007, Blankertz et al. 2004, 2005, 2006, 2007, 2008]

BBCI Set-up

Artifact removal

[cf. Mller et al. 2001, 2007, 2008, Dornhege et al. 2003, 2007, Blankertz et al. 2004, 2005, 2006, 2007, 2008]

36

Slide: What can Machine Learning tell us about physiology?

1

1

[cf. Blankertz et al. 2001, 2006]

What can Machine Learning tell us about physiology?

1

1

[cf. Blankertz et al. 2001, 2006]

37

Slide: BCI with machine learning: feedback

BCI with machine learning: feedback

38

Slide: Spelling with BBCI: a communication for the disabled I

Spelling with BBCI: a communication for the disabled I

39

Slide: Spelling with BBCI: a communication for the disabled II

Spelling with BBCI: a communication for the disabled II

40

Slide: Variance IV: Shifting distributions within experiment

Variance IV: Shifting distributions within experiment

41

Slide: Interlude: Caveats in Validation

[cf. Blankertz et al 2011]

Interlude: Caveats in Validation

[cf. Blankertz et al 2011]

42

Slide: Hall of pitfalls in single-trial EEG analysis (and beyond)

Hall of pitfalls in single-trial EEG analysis (and beyond)

43

Slide: Block design

Block design

44

Slide: Slowly varying variables

Slowly varying variables

45

Slide: A validation test

A validation test

46

Slide: Results of validation test

Results of validation test

47

Slide: Further remarks & summary

Further remarks & summary

48

Slide: Part II ML challenges

 Aleviating non-stationarity  Multimodal sources

Part II ML challenges

Aleviating non-stationarity Multimodal sources

49

Slide: Recap: BBCI Set-up

Artifact removal

[cf. Mller et al. 2001, 2007, 2008, Dornhege et al. 2003, 2007, Blankertz et al. 2004, 2005, 2006, 2007, 2008]

Recap: BBCI Set-up

Artifact removal

[cf. Mller et al. 2001, 2007, 2008, Dornhege et al. 2003, 2007, Blankertz et al. 2004, 2005, 2006, 2007, 2008]

50

Slide: Nonstationarity in BCI

Nonstationarity in BCI

51

Slide: Variance IV: Shifting distributions within experiment

Variance IV: Shifting distributions within experiment

52

Slide: Mathematical flavors of non-stationarity

- Bias adaptation between training and test

- Covariate shift

- SSA: projecting to stationary subspaces

- Nonstationarity due to subject dependence: Mixed effects model

- Co-adaptation

Mathematical flavors of non-stationarity

- Bias adaptation between training and test

- Covariate shift

- SSA: projecting to stationary subspaces

- Nonstationarity due to subject dependence: Mixed effects model

- Co-adaptation

53

Slide: Neurophysiological analysis

[cf. Krauledat et al. 07]

Neurophysiological analysis

[cf. Krauledat et al. 07]

54

Slide: Weighted Linear Regression for covariate shift compensation

, choosing

yields unbiased estimator even under covariate shift

[cf. Sugiyama & Mller 2005, Sugiyama et al. JMLR 2007, see next week MLSS12]

Weighted Linear Regression for covariate shift compensation

, choosing

yields unbiased estimator even under covariate shift

[cf. Sugiyama & Mller 2005, Sugiyama et al. JMLR 2007, see next week MLSS12]

55

Slide: Projection Methods: recap

Projection Methods: recap

56

Slide: Splitting into stationary and nonstationary subspace: SSA

invert
[cf. Bnau, Meinecke, Kiraly, Mller PRL 09]

Splitting into stationary and nonstationary subspace: SSA

invert
[cf. Bnau, Meinecke, Kiraly, Mller PRL 09]

57

Slide: SSA

SSA

58

Slide: Inverting the SSA Mixing Model

Inverting the SSA Mixing Model

59

Slide: SSA: Algorithm idea

SSA: Algorithm idea

60

Slide: Algorithm idea

Algorithm idea

61

Slide: Using Symmetries and Invariances

Using Symmetries and Invariances

62

Slide: SSA: Objective Function

SSA: Objective Function

63

Slide: Optimzing

Optimzing

64

Slide: Spurious stationarity

Spurious stationarity

65

Slide: SSA: how many epochs?

SSA: how many epochs?

66

Slide: How many epochs? Theoretical results

How many epochs? Theoretical results

67

Slide: Simulations: toy data

Simulations: toy data

68

Slide: Application to Brain-Computer-Interfacing

Application to Brain-Computer-Interfacing

69

Slide: Real Man Machine Interaction

Real Man Machine Interaction

70

Slide: Towards a subject independent BCI decoder

Towards a subject independent BCI decoder

71

Slide: Model formulation

Model formulation

72

Slide: Linear Mixed Effects Model: intuition

[Fazli, Mller et al. 2011]

Linear Mixed Effects Model: intuition

[Fazli, Mller et al. 2011]

73

Slide: Approach to Cure BCI Illiteracy

Runs 4 6

supervised CSP + sel.Lap. unsupervised CSP

Runs 1 3

fixed Laplace

 Direct feedback -> Unspecific LDA classifier.  Each trial, perform adaptation of the cls.  Features: log band power (alpha and beta).  Laplacian channels C3, C4 and Cz.

 Compute CSP and sel. Laps. from runs 1-3.  Fixed CSP filters, automated laps. selection.  Each trial retrain the classifier.

Runs 7 8

 Compute CSP from runs 4-6.  Perform unsupervised adaptation of pooled mean.  Update the bias of the classifier.
[cf. Vidaurre, Blankertz, Mller et al. Neural Comp. to appear]

Approach to Cure BCI Illiteracy

Runs 4 6

supervised CSP + sel.Lap. unsupervised CSP

Runs 1 3

fixed Laplace

Direct feedback -> Unspecific LDA classifier. Each trial, perform adaptation of the cls. Features: log band power (alpha and beta). Laplacian channels C3, C4 and Cz.

Compute CSP and sel. Laps. from runs 1-3. Fixed CSP filters, automated laps. selection. Each trial retrain the classifier.

Runs 7 8

Compute CSP from runs 4-6. Perform unsupervised adaptation of pooled mean. Update the bias of the classifier.
[cf. Vidaurre, Blankertz, Mller et al. Neural Comp. to appear]

74

Slide: Results (Grand Averages)

Results (Grand Averages)

75

Slide: Example: one subject of Cat. III

Runs 1 and 2

!

Runs 7 and 8

[cf. Vidaurre, Blankertz, Mller et al. 2009]

Example: one subject of Cat. III

Runs 1 and 2

!

Runs 7 and 8

[cf. Vidaurre, Blankertz, Mller et al. 2009]

76

Slide: Multimodal

Multimodal

77

Slide: Different physiological Features
EEG signals

 

Slow Features, e.g. Event Related Potential/Slow Cortical Potentials (ERP/SCP)
Independent??? Neurophysiology: YES

Maps

 

Oscillatory Features, e.g. Event Related Desynchronization/ Synchronization (ERD/ERS)


[Dornhege, et al. 2006]

Different physiological Features
EEG signals

Slow Features, e.g. Event Related Potential/Slow Cortical Potentials (ERP/SCP)
Independent??? Neurophysiology: YES

Maps

Oscillatory Features, e.g. Event Related Desynchronization/ Synchronization (ERD/ERS)

[Dornhege, et al. 2006]

78

Slide: Different physiological Features

Different physiological Features

79

Slide: Independent Features
Covariance matrix between features Distribution of misclassified and classified trials for different features (loo) Correlation of classifier output (continuous/ label)

from left to right, top to bottom: MRP, AR,CSP

Independent Features
Covariance matrix between features Distribution of misclassified and classified trials for different features (loo) Correlation of classifier output (continuous/ label)

from left to right, top to bottom: MRP, AR,CSP

80

Slide: Combination Results

Combination Results

81

Slide: Combination Results

The figures show the Information Transfer Rate per decision for the best single feature compared to the suggested algorithms on all subset of classes out of the experiments we have done. Above each figure a histogram is plotted. For points right of the middle line the suggested algorithm outperforms the best single feature performance.

Combination Results

The figures show the Information Transfer Rate per decision for the best single feature compared to the suggested algorithms on all subset of classes out of the experiments we have done. Above each figure a histogram is plotted. For points right of the middle line the suggested algorithm outperforms the best single feature performance.

82

Slide: Example: NIRS-EEG Brain Computer Interfaces

[Fazli et al. Neuroimage 2012]

94

Example: NIRS-EEG Brain Computer Interfaces

[Fazli et al. Neuroimage 2012]

94

83

Slide: Photon Transport in the Human Brain Tissue

 Near-Infrared light can penetrate the brain  banana-shaped measurement volume for non-invasive NIRS
95

Photon Transport in the Human Brain Tissue

Near-Infrared light can penetrate the brain banana-shaped measurement volume for non-invasive NIRS
95

84

Slide: Experimental Setup and Paradigm
EEG: 37 electrodes NIRS 26 channels (frontal, parietal, occipital) EEG-based cursor feedback (ISI = 15 s) Executed movement vs imagery movements Imagery movements: EEG-feedback for left and right motor imagery

Number of subjects: 14

Can a simultaneous measurement of NIRS and EEG during Brain Computer Interfacing enhance the classification accuracy? Are the results physiologically reliable?

Fazli et al. 2012

96

Experimental Setup and Paradigm
EEG: 37 electrodes NIRS 26 channels (frontal, parietal, occipital) EEG-based cursor feedback (ISI = 15 s) Executed movement vs imagery movements Imagery movements: EEG-feedback for left and right motor imagery

Number of subjects: 14

Can a simultaneous measurement of NIRS and EEG during Brain Computer Interfacing enhance the classification accuracy? Are the results physiologically reliable?

Fazli et al. 2012

96

85

Slide: Temporal Dependency of Classification in Executed Movements

EEG

HbO

HbR

Fazli et al. 2012

EEG peaks earlier as compared to HbO and HbR Physiological reliability: HRF shaped classification accuracies over time Classification accuracy higher for EEG

97

Temporal Dependency of Classification in Executed Movements

EEG

HbO

HbR

Fazli et al. 2012

EEG peaks earlier as compared to HbO and HbR Physiological reliability: HRF shaped classification accuracies over time Classification accuracy higher for EEG

97

86

Slide: Temporal Dependency of Classification in Motor Imagery

EEG peaks earlier as compared to HbO and HbR Physiological reliability: HRF shaped classification accuracies over time Classification accuracy higher for EEG Classification accuracy lower than in executed movements

98

Temporal Dependency of Classification in Motor Imagery

EEG peaks earlier as compared to HbO and HbR Physiological reliability: HRF shaped classification accuracies over time Classification accuracy higher for EEG Classification accuracy lower than in executed movements

98

87

Slide: Topography for Executed Movements

EEG

HbO

HbR

EEG earlier NIRS has clear lateralization HbO goes up, HbR down

99

Topography for Executed Movements

EEG

HbO

HbR

EEG earlier NIRS has clear lateralization HbO goes up, HbR down

99

88

Slide: Topography for Imagery Movements

EEG

HbO

HbR

Similar results EEG earlier NIRS has clear lateralization HbO goes up, HbR up (reason unsolved)

100

Topography for Imagery Movements

EEG

HbO

HbR

Similar results EEG earlier NIRS has clear lateralization HbO goes up, HbR up (reason unsolved)

100

89

Slide: Combination of EEG and NIRS

Fazli et al. 2012

LDA classifier estimated for EEG, HbO and HbR (individually) Meta-classifier estimated for combination in each subject All within cross-validation (8 chronological splits)

101

Combination of EEG and NIRS

Fazli et al. 2012

LDA classifier estimated for EEG, HbO and HbR (individually) Meta-classifier estimated for combination in each subject All within cross-validation (8 chronological splits)

101

90

Slide: Feature Combination
Fazli et al. 2012

NIRS-EEG combinations have higher classification accuracies for vast majority of subjects

102

Feature Combination
Fazli et al. 2012

NIRS-EEG combinations have higher classification accuracies for vast majority of subjects

102

91

Slide: Feature Combination

t-tests reveal a significant increase of classification accuracy for combination

Fazli et al. 2012

103

Feature Combination

t-tests reveal a significant increase of classification accuracy for combination

Fazli et al. 2012

103

92

Slide: Feature Combination

Some subjects, which were not classifiable with EEG become classifiable by a metaclassifier in combination with NIRS

104

Feature Combination

Some subjects, which were not classifiable with EEG become classifiable by a metaclassifier in combination with NIRS

104

93

Slide: Mutual Information

NIRS features for all correct EEG trials (EEG+) and incorrect EEG trials (EEG-) Pattern is similar although the significance drops NIRS can complement the EEG with physiological meaningful information

105

Mutual Information

NIRS features for all correct EEG trials (EEG+) and incorrect EEG trials (EEG-) Pattern is similar although the significance drops NIRS can complement the EEG with physiological meaningful information

105

94

Slide: Discussion
Problems


 

Different temporal properties of the measurement devices (e.g. EEG: 1000 Hz, NIRS: max. 10 Hz)
Temporal lag between parameters Different signal qualities

Ideas to Overcome the Temporal Lag    NIRS as a measure of subjects attention to predict EEG-based performance NIRS as a localizer of the source of EEG signals NIRS as a stop, e.g. to discard a EEG-based classified trial when not confirmed by NIRS

106

Discussion
Problems

Different temporal properties of the measurement devices (e.g. EEG: 1000 Hz, NIRS: max. 10 Hz)
Temporal lag between parameters Different signal qualities

Ideas to Overcome the Temporal Lag NIRS as a measure of subjects attention to predict EEG-based performance NIRS as a localizer of the source of EEG signals NIRS as a stop, e.g. to discard a EEG-based classified trial when not confirmed by NIRS

106

95

Slide: Correlating apples and oranges

[Biessmann et al. Neuroimage 2012, Machine Learning 2010]

107

Correlating apples and oranges

[Biessmann et al. Neuroimage 2012, Machine Learning 2010]

107

96

Slide: CCA: correlating apples and oranges

CCA: correlating apples and oranges

97

Slide: kCCA: solving CCA on data kernels

kCCA: solving CCA on data kernels

98

Slide: tkCCA: correlating apples and oranges over time

tkCCA: correlating apples and oranges over time

99

Slide: CCA: correlating apples and oranges

CCA: correlating apples and oranges

100

Slide: Experimental Setup

Experimental Setup

101

Slide: Temporal Kernel CCA

Temporal Kernel CCA

102

Slide: Results tkCCA: spatial dependencies and HRF

Results tkCCA: spatial dependencies and HRF

103

Slide: Conclusion II

FOR INFORMATION SEE: www.bbci.de

Conclusion II

FOR INFORMATION SEE: www.bbci.de

104

Slide: Part III ERP analysis & applications beyond communication

Part III ERP analysis & applications beyond communication

105

Slide: Neurophysiological Background for ERPs

Neurophysiological Background for ERPs

106

Slide: Experimental Design

Experimental Design

107

Slide: P300 in action: Hex-o-spell

P300 in action: Hex-o-spell

108

Slide: Single subject ERPs for Hex-o-spell

Single subject ERPs for Hex-o-spell

109

Slide: Topographies of ERP components

Topographies of ERP components

110

Slide: Classification of temporal features

Classification of temporal features

111

Slide: Extraction of spatial features

Extraction of spatial features

112

Slide: The r^2 matrix of differences

The r^2 matrix of differences

113

Slide: Spatial features

Spatial features

114

Slide: A linear classifier as a spatial filter

A linear classifier as a spatial filter

115

Slide: Classification results of spatial features

Classification results of spatial features

116

Slide: Extraction of spatio-temporal features

Extraction of spatio-temporal features

117

Slide: Spatio-temporal features

Spatio-temporal features

118

Slide: Classification results for spatio-temporal features

Classification results for spatio-temporal features

119

Slide: Bias in estimating covariances

Bias in estimating covariances

120

Slide: Bias in estimating covariances II

Bias in estimating covariances II

121

Slide: A remedy for classification

A remedy for classification

122

Slide: Modelselection

Modelselection

123

Slide: Regularized LDA at work

Regularized LDA at work

124

Slide: Investigating the impact of shrinkage

Investigating the impact of shrinkage

125

Slide: ERP and noise

ERP and noise

126

Slide: Spatial structure of noise

Spatial structure of noise

127

Slide: Understanding spatial filters

Understanding spatial filters

128

Slide: Understanding spatial filters II

Understanding spatial filters II

129

Slide: Impact of shrinkage on the spatial filters

Impact of shrinkage on the spatial filters

130

Slide: Optimal selection of shrinkage parameters

Optimal selection of shrinkage parameters

131

Slide: Result of Classification with shrinkage

Result of Classification with shrinkage

132

Slide: Summary spatio-temporal classification

Summary spatio-temporal classification

133

Slide: Applications

Applications

134

Slide: Clinical Applications

Clinical Applications

135

Slide: Towards industrial applications of BCI Technology

Technology

[Blankertz et al 2010 Front. Neurosci.]

Towards industrial applications of BCI Technology

Technology

[Blankertz et al 2010 Front. Neurosci.]

136

Slide: Operant conditioning: Tbingen Group

Operant conditioning: Tbingen Group

137

Slide: Non-Invasive: Tbingen. Birbaumer Lab: Slow Cortical potentials
Negativity task

Positivity task

Amplitude [V]

-15 -10 -5 0 5 10 15

negativity task positivity task

0

1

2

3

4

5

6

7

8

Time [s]

[From Birbaumer et al.]

Non-Invasive: Tbingen. Birbaumer Lab: Slow Cortical potentials
Negativity task

Positivity task

Amplitude [V]

-15 -10 -5 0 5 10 15

negativity task positivity task

0

1

2

3

4

5

6

7

8

Time [s]

[From Birbaumer et al.]

138

Slide: [From Birbaumer et al.]

SCP

[From Birbaumer et al.]

SCP

139

Slide: [From Birbaumer et al.]

[From Birbaumer et al.]

140

Slide: [From Birbaumer et al.]

[From Birbaumer et al.]

141

Slide: ECOG Decoding

ECOG Decoding

142

Slide: ECOG

 presurgical localization of area causing epilepsy

 excellent possibilty to learn about brain for human subject

[From Schalk]

ECOG

presurgical localization of area causing epilepsy

excellent possibilty to learn about brain for human subject

[From Schalk]

143

Slide: Index vs rest Thumb vs rest

[From Schalk]

Index vs rest Thumb vs rest

[From Schalk]

144

Slide: ECOG Analysis

[From Schalk]

ECOG Analysis

[From Schalk]

145

Slide: fMRI Decoding

fMRI Decoding

146

Slide: Example: Which Video are you watching?
 Study: Reconstructing Visual Experience from Brain Activity Evoked by Natural Movies (Nishimoto 2011)  Aim: validation of neurovascular coupling in the visual cortex

 Models of hemodynamics elicited by a movie for each voxel in early visual areas  fMRI measurement of subjects watching movies  Reconstruction of movies from the brains activity

Example: Which Video are you watching?
Study: Reconstructing Visual Experience from Brain Activity Evoked by Natural Movies (Nishimoto 2011) Aim: validation of neurovascular coupling in the visual cortex

Models of hemodynamics elicited by a movie for each voxel in early visual areas fMRI measurement of subjects watching movies Reconstruction of movies from the brains activity

147

Slide:

148

Slide: Example: Which Video are you watching?

Motion energy of the pictures were calculated and fed to hemodynamic modeling

Example: Which Video are you watching?

Motion energy of the pictures were calculated and fed to hemodynamic modeling

149

Slide: Example: Which Video are you watching?
 Bayesian fit to acquired data of 3 subjects watching 12 movie (each once)

 Test the approach on subject watching 9 other movies (each 10 times)

Example: Which Video are you watching?
Bayesian fit to acquired data of 3 subjects watching 12 movie (each once)

Test the approach on subject watching 9 other movies (each 10 times)

150

Slide: Example: Which Video are you watching?

The accuracy becomes worse when more films are included for decoding (not watched by the subjects) but remains high

Example: Which Video are you watching?

The accuracy becomes worse when more films are included for decoding (not watched by the subjects) but remains high

151

Slide: Towards industrial applications of BCI Technology

Technology

[Blankertz et al 2010 Front. Neurosci.]

Towards industrial applications of BCI Technology

Technology

[Blankertz et al 2010 Front. Neurosci.]

152

Slide: BCI for Assessing Signal Quality perception

BCI for Assessing Signal Quality perception

153

Slide: Why Quality Assessment?
  Ensure user satisfaction Develop better compression algorithms

quality assessment profitability signal quality

 www.eftrends.com

Why Quality Assessment?
Ensure user satisfaction Develop better compression algorithms

quality assessment profitability signal quality

www.eftrends.com

154

Slide: Approaches
Behavioral tests (standard) EEG + BCI methods (novel)

 Continuous signal  Objective measure  Capture > subtle differences > non-conscious processing

Approaches
Behavioral tests (standard) EEG + BCI methods (novel)

Continuous signal Objective measure Capture > subtle differences > non-conscious processing

155

Slide: EEG Studies

Domain Auditory Visual

Stimuli Phonemes

Cooperation Partner Telekom Laboratories

Words
Flickering light Video

-Philips Research Fraunhofer (HHI)

EEG Studies

Domain Auditory Visual

Stimuli Phonemes

Cooperation Partner Telekom Laboratories

Words
Flickering light Video

-Philips Research Fraunhofer (HHI)

156

Slide: Audio Quality
   Discrimination task: Is stimulus disturbed? Recording: button press, 64-channel EEG Stimuli:  4 levels of degradation: strong (T1)  weak (T4),  undisturbed stimulus (NT)

Phoneme Study stimulus disturbed by /a/ signal-correlated noise

Word Study /Haus/, /Schild/ by female/male speaker bit rate limitation

Audio Quality
Discrimination task: Is stimulus disturbed? Recording: button press, 64-channel EEG Stimuli: 4 levels of degradation: strong (T1) weak (T4), undisturbed stimulus (NT)

Phoneme Study stimulus disturbed by /a/ signal-correlated noise

Word Study /Haus/, /Schild/ by female/male speaker bit rate limitation

157

Slide: Audio Quality
 Hits: The more subtle the noise, the lower the amplitude and the higher the latency of P3 component  Neural effort  Quantification of hits

Grand average EEG signal (ERP): stimulus T1 (strong degradation), T3 (weak degradation), NT (undisturbed).

Audio Quality
Hits: The more subtle the noise, the lower the amplitude and the higher the latency of P3 component Neural effort Quantification of hits

Grand average EEG signal (ERP): stimulus T1 (strong degradation), T3 (weak degradation), NT (undisturbed).

158

Slide: Audio Quality

Phonemes

Words

Audio Quality

Phonemes

Words

159

Slide: Audio Quality
 Misses: Similarity to hits at the threshold of perception  Non-conscious processing  Quantified by linear classification

Difference topographies at the threshold of perception: hits / misses (low quality)  correct rejections (high quality) (one participant, phonemes)

Audio Quality
Misses: Similarity to hits at the threshold of perception Non-conscious processing Quantified by linear classification

Difference topographies at the threshold of perception: hits / misses (low quality) correct rejections (high quality) (one participant, phonemes)

160

Slide: Visual Quality
   Discrimination task: Does the stimulus flicker? Recording: 64-channel EEG, button press Stimuli:  Constant wave light (CW)  4 levels of flicker frequency: slow (S1)  fast (S4)
LED light source red diode

Visual Quality
Discrimination task: Does the stimulus flicker? Recording: 64-channel EEG, button press Stimuli: Constant wave light (CW) 4 levels of flicker frequency: slow (S1) fast (S4)
LED light source red diode

161

Slide: Visual Quality
 Added value of EEG

Stimulation frequencies [Hz] per participant; colored cells: significant neural response
- Orange: shown by EEG (t-test, univariate)

Visual Quality
Added value of EEG

Stimulation frequencies [Hz] per participant; colored cells: significant neural response
- Orange: shown by EEG (t-test, univariate)

162

Slide: Visual Quality
 Added value of EEG and ML

Stimulation frequencies [Hz] per participant; colored cells: significant neural response
- Yellow + orange: shown by ML (CSP+LDA, multivariate)

Visual Quality
Added value of EEG and ML

Stimulation frequencies [Hz] per participant; colored cells: significant neural response
- Yellow + orange: shown by ML (CSP+LDA, multivariate)

163

Slide: Visual Quality Gain by NT
  Discrimination task: Does the stimulus flicker? LED light source Stimuli: slow (S1)  fast (S4) flfr & CW
red diode

Visual Quality Gain by NT
Discrimination task: Does the stimulus flicker? LED light source Stimuli: slow (S1) fast (S4) flfr & CW
red diode

164

Slide: Video Quality
  Detection task: Does the quality change in the video? Stimuli:  artificially generated videos (8 sec) with a quality change  Undistorted baseline (BL), 8 levels of distortion (S1-8) Recording: 64-channel EEG, button press

Video Quality
Detection task: Does the quality change in the video? Stimuli: artificially generated videos (8 sec) with a quality change Undistorted baseline (BL), 8 levels of distortion (S1-8) Recording: 64-channel EEG, button press

165

Slide: Video Quality

Video Quality

166

Slide: Video Quality
 P3 component is a graded neural index of quality perception (left) Effect depends on subjective perception (right) Non-conscious processing in 3 out of 11 participants

Video Quality
P3 component is a graded neural index of quality perception (left) Effect depends on subjective perception (right) Non-conscious processing in 3 out of 11 participants

167

Slide: Summary
 Audio Quality Neuronal effort: loss of quality is reflected in P3 latency/amplitude

-

Non-Conscious Processing. use classification to single out trials where misses resemble hits
Visual Quality Non-Conscious Processing: high-frequency flicker can still elicit a neural response, even if it is not noticed behaviorally

 -

-

Machine Learning: classification reveals effect for additional participants and stimuli

Summary
Audio Quality Neuronal effort: loss of quality is reflected in P3 latency/amplitude

-

Non-Conscious Processing. use classification to single out trials where misses resemble hits
Visual Quality Non-Conscious Processing: high-frequency flicker can still elicit a neural response, even if it is not noticed behaviorally

-

-

Machine Learning: classification reveals effect for additional participants and stimuli

168

Slide: BCI for Assessing Workload

BCI for Assessing Workload

169

Slide: Nonclinical Application: tiredness monitoring

[Kohlmorgen, Mller et al 2007]

Nonclinical Application: tiredness monitoring

[Kohlmorgen, Mller et al 2007]

170

Slide: Application: Cognitive workload and drowsyness assessment
Assess workload with BCI and balance it by smart driver assistent system
Assess cognitive alertness

[Kohlmorgen, Mller et al 2007]

Application: Cognitive workload and drowsyness assessment
Assess workload with BCI and balance it by smart driver assistent system
Assess cognitive alertness

[Kohlmorgen, Mller et al 2007]

171

Slide: BCI for Assessing Upcoming decisions

BCI for Assessing Upcoming decisions

172

Slide: Bereitschaftspotential over C3 (primary motor cortex of the right hand)

V -1.5 -1.0 -0.5 0

reactive spontaneous

- 0.60 s

- 0.25 s

movement right hand

Bereitschaftspotential over C3 (primary motor cortex of the right hand)

V -1.5 -1.0 -0.5 0

reactive spontaneous

- 0.60 s

- 0.25 s

movement right hand

173

Slide: EEG single-trial preprocessing

EEG single-trial preprocessing

174

Slide: Regularized Fisher Discriminant

Regularized Fisher Discriminant

175

Slide: Fishers Discriminant: Assumptions correct?

Fishers Discriminant: Assumptions correct?

176

Slide: Linear Classification
V at C4 hyperplane
1) Binary linear classification separates the feature space by a hyperplane. 2) Projection line: 'best' discriminating dimension 3) Linear classifications determine a projection line on a training set such that a specific objective is satisfied for the projected distributions.

projection line

V at C3

E.g. Fisher Discriminant (FD) maximizes margin between means of the projected class distributions and minimizes intra-class variance.

 Linear classifications yields good generalisation in case of limited training data.
 BUT Regularize!

Linear Classification
V at C4 hyperplane
1) Binary linear classification separates the feature space by a hyperplane. 2) Projection line: 'best' discriminating dimension 3) Linear classifications determine a projection line on a training set such that a specific objective is satisfied for the projected distributions.

projection line

V at C3

E.g. Fisher Discriminant (FD) maximizes margin between means of the projected class distributions and minimizes intra-class variance.

Linear classifications yields good generalisation in case of limited training data.
BUT Regularize!

177

Slide: Robustness against outliers is mandatory

Robustness against outliers is mandatory

178

Slide: Time development of classification error (FDA)
At an average keystroke interval of 2.1 sec  22.9 bit/min

endpoint of classification window
EEG error  10% after -230 ms before keystroke

keystroke

keystroke

Time development of classification error (FDA)
At an average keystroke interval of 2.1 sec 22.9 bit/min

endpoint of classification window
EEG error 10% after -230 ms before keystroke

keystroke

keystroke

179

Slide: Steps towards online classification
 no usage of information about event timing (keystrokes)  ternary decision: right  left  no movement
 continuous classification in sliding windows + graded output detect upcoming movements predict movement laterality

online 2-classifier combination: 10% error rate corresponding to 29 bits/min.

Steps towards online classification
no usage of information about event timing (keystrokes) ternary decision: right left no movement
continuous classification in sliding windows + graded output detect upcoming movements predict movement laterality

online 2-classifier combination: 10% error rate corresponding to 29 bits/min.

180

Slide:

181

Slide:

182

Slide: The shape of thoughts to come

LEFT hand

RIGHT hand

The shape of thoughts to come

LEFT hand

RIGHT hand

183

Slide: Study: emergency breaking in driving simulator

- Highly specific sequence of EEG potentials 500 ms before breaking 1) Perception of breaklight stimulus (visual evoked potentials) 2) Identification of emergency (P300 component) 3) Preparation of breaking movement (Bereitschaftspotential) - EEG (+EMG) features improve the pedal based breaking detector by 150 ms

- 4 m less breaking space at speed100 km/h

[Haufe et al., EEG potentials predict upcoming emergency brakings during simulated driving. J Neural Eng. 2011]

Study: emergency breaking in driving simulator

- Highly specific sequence of EEG potentials 500 ms before breaking 1) Perception of breaklight stimulus (visual evoked potentials) 2) Identification of emergency (P300 component) 3) Preparation of breaking movement (Bereitschaftspotential) - EEG (+EMG) features improve the pedal based breaking detector by 150 ms

- 4 m less breaking space at speed100 km/h

[Haufe et al., EEG potentials predict upcoming emergency brakings during simulated driving. J Neural Eng. 2011]

184

Slide:

185

Slide: Car Safety: Improving emergency braking

Car Safety: Improving emergency braking

186

Slide: Conclusion

 BBCI: Untrained, Calibration < 10min, data analysis <<5min, BCI experiment
 5-8 letters/min mental typewriter CeBit 06,10. Brain2Robot@Medica 07, lNdW 09  Machine Learning and modern data analysis is of central importance for BCI et al  Important issue of this talk: How to learn under nonstationarity?  Solutions:  SSA, i.e. project on stationary subspace and learn there, linear, sound & fast  Modeling: covariate shift based CV: special  mixed effects model  co-adaptation, Multimodal  tracking, invariant features etc

FOR INFORMATION SEE: www.bbci.de

Conclusion

BBCI: Untrained, Calibration < 10min, data analysis <<5min, BCI experiment
5-8 letters/min mental typewriter CeBit 06,10. Brain2Robot@Medica 07, lNdW 09 Machine Learning and modern data analysis is of central importance for BCI et al Important issue of this talk: How to learn under nonstationarity? Solutions: SSA, i.e. project on stationary subspace and learn there, linear, sound & fast Modeling: covariate shift based CV: special mixed effects model co-adaptation, Multimodal tracking, invariant features etc

FOR INFORMATION SEE: www.bbci.de

187

Slide: Before-after Future issues: sensors

Popescu et al 2007

Before-after Future issues: sensors

Popescu et al 2007

188

Slide: Before-after

Before-after

189

Slide:

190

Slide: Thanks to BBCI core team:
Gabriel Curio Florian Losch Volker Kunzmann Frederike Holefeld Vadim Nikulin@Charite

Florin Popescu Andreas Ziehe Steven Lemm Motoaki Kawanabe Guido Nolte@FIRST Yakob Badower@Pico Imaging Marton Danozci

Benjamin Blankertz Michael Tangermann Claudia Sannelli Carmen Vidaurre Siamac Fazli Martijn Schreuter Stefan Haufe Laura Acqualagna Thorsten Dickhaus Frank Meinecke Felix Biessmann@TUB
Matthias Krauledat Guido Dornhege Roman Krepki@industry

Collaboration with: U Tbingen, Bremen, Albany, TU Graz, EPFL, Daimler, Siemens, MES, MPIs, U Tokyo, TIT, RIKEN, Bernstein Center for Computational Neuroscience Berlin, Columbia, CUNY Funding by: EU, BMBF and DFG

Thanks to BBCI core team:
Gabriel Curio Florian Losch Volker Kunzmann Frederike Holefeld Vadim Nikulin@Charite

Florin Popescu Andreas Ziehe Steven Lemm Motoaki Kawanabe Guido Nolte@FIRST Yakob Badower@Pico Imaging Marton Danozci

Benjamin Blankertz Michael Tangermann Claudia Sannelli Carmen Vidaurre Siamac Fazli Martijn Schreuter Stefan Haufe Laura Acqualagna Thorsten Dickhaus Frank Meinecke Felix Biessmann@TUB
Matthias Krauledat Guido Dornhege Roman Krepki@industry

Collaboration with: U Tbingen, Bremen, Albany, TU Graz, EPFL, Daimler, Siemens, MES, MPIs, U Tokyo, TIT, RIKEN, Bernstein Center for Computational Neuroscience Berlin, Columbia, CUNY Funding by: EU, BMBF and DFG

191

Slide: BCI Competitions

For BCI IV Competition see www.bbci.de

BCI Competitions

For BCI IV Competition see www.bbci.de

192

Slide: FOR INFORMATION SEE: www.bbci.de

Machine Learning open source software initiative: MLOSS see www.jmlr.org

FOR INFORMATION SEE: www.bbci.de

Machine Learning open source software initiative: MLOSS see www.jmlr.org

193

Slide:

194

Slide:

195

Slide:

196

Slide: Biased selected references

Biased selected references

197