|
Class Related:
Course Syllabus: This syllabus
applies to the CIS 8390 course that
Dr. Allison is teaching during summer
2004 at GSU. Syllabi for the BCI class
taught last year at UCSD are available
upon request.
Wolpaw et al. (1991): This is the
first published paper about mu BCIs.
It is required for week 2.
Wolpaw et al (1994): This is the
first published paper about 2
dimensional mu BCIs. It is required
for week 2.
Wolpaw et al. (2002): This is the
only published review of BCIs. It
describes the key parts of any BCI
(operating system including
application, signal acquisition,
feature extraction, translation
algorithm) and the five types of BCIs
(P300, mu, SSVEP, SCP, implanted).
This is required reading for week 1.
Wickelgren (2003): This is a short
article from Science magazine
describing some different types of
BCIs and their application as
communication systems for locked in
users. It is required for week 1.
Allison PhD thesis (2003): Chapter
2 of this thesis is required for week
1. This chapter provides an
introduction to how EEGs work and how
they can be applied to BCIs. You do
not need to memorize details such as
the minutiae of section 2.5.1, but
should be familiar with different
types of EEG activity that could be
used in BCIs.
Schalk et al. (2004): This is
required for week 1. It presents
BCI2000, the software platform used by
many BCI labs.
Wolpaw et al. (2003):
This is a
brief review of the research program
at the Wolpaw lab as of last year. It
presents cutting edge research in mu
BCIs. It is required for week 2.
Pfurtscheller et al. (2003): This
is a brief review of the research
program at the Pfurtscheller lab as of
2003. It presents cutting edge
research in mu BCIs. It is required for week 2.
Blankertz et al. (2003):
This
presents a recently held BCI data
analysis competition. It shows several
types of BCI data and which pattern
recognition approaches were most
effective. It is required for week 2.
Kostav and Polak (2000): This very
short article is also required for
week 2. It presents a somewhat
different type of 2D mu BCI.
Scherer et al (2004): This is a
short article from the Pfurtscheller
lab. It presents a virtual keyboard
controlled by 2D mu activity. It is
required for week 2.
Birch et al (2003): This is
required for week 2. It presents an
overview of research at the Neil
Squire Foundation, which has a mu
based switch.
*** ADDITIONAL READINGS FOR WEEK 2
WERE HANDED OUT IN CLASS, AS
ELECTRONIC COPIES ARE NOT AVAILABLE.
THIS WAS KALCHER ET AL. (1996). ***
Obermaier et al. (2003): The above
paper (Scherer et al 2004) is an
extension of this paper, which
presents a menu based speller rather
than a virtual keyboard. It is
recommended for week 2.
Muller et al. (2003): This
demonstrates a telemonitoring system
that allows researchers to modify
various BCI parameters from remote
locations. It is recommended for week
2.
McFarland et al. (2003): This is
recommended for
week 2. It presents issues in
optimizing information transfer rates,
and the results recap an important
point: different subjects often work
best with different parameters.
Week 3 readings:
Farwell and Donchin (1988): This
is required for week 3. It is the
first paper about P300 BCIs.
Donchin et al. (2000): This is
required for week 3. It is a follow up
to the Farwell and Donchin '88 paper
handed out in class. It explores the
use of a discrete wavelet transform (DWT).
Bayliss and Ballard (2000): This
is required for week 3. It presents a
P300 BCI used in a virtual driving
simulator.
Bayliss (2003): This is required
for week 3. It presents a P300 BCI for
selecting icons that could be used to
send various commands in an apartment
setting.
Allison and Pineda (2003): This is
required for week 3. It shows the
relationship between set size and ERP
measures, performance, and user
factors.
Kaper et al (2004): This is
required for week 3. It describes a
pattern classification approach called
an SVM, or support vector machine,
which is a type of neural network. It
did well with P300 data.
Xu et al. (2004): This is required
for week 3. It describes an ICA based
approach to categorizing P300 data.
Green and Bavelier (2003): This is
recommended for week 3. It doesn't
have P300s or any EEGs, but suggests
that video game players would have
different EEGs than healthy subjects.
This was also found in Allison's PhD
thesis.
Allison et al. (2002): This is a
conference poster presenting the first
and second studies of Allison's
thesis. The second study presents the
same material as Allison and Pineda
(2003). It is recommended for week 3.
Middendorf et al. (2000): This is
required for week 3. It presents a
monitor-based SSVEP system and a few
cool applications for it.
Gao et al. (2003): This is
required for week 3. It describes an
SSVER apparatus with 48 LEDs.
Gao et al. (2002): This is
recommended for week 3. In it,
subjects use a monitor based SSVEP
mechanism to send a phone number and
spell.
*** ADDITIONAL READINGS FOR WEEK 3
WERE HANDED OUT IN CLASS, AS
ELECTRONIC COPIES ARE NOT AVAILABLE.
THESE WERE VIDAL (1977), SUTTER AND TRAN
(1992), AND TWO 1995 RESNA PAPERS FROM
MACMILLAN'S GROUP ***
Week 4 readings:
Birbaumer et al. (1999): This is
required for week 4. The famous Nature
paper shows a letter written by a
locked in patient using the TTD.
Birbaumer et al. (2003): This presents
an overview of the Birbaumer lab's BCI
(called the TTD, or thought
translation device) as of 2003. It is
required for week 4.
Hinterberger et al (2004): This
paper lays out the theoretical basis
for a combined fMRI/EEG based BCI. It
is required for week 4.
Schalk et al (2000): This paper
shows the presence of error activity
in BCI users' brain activity after a
mistake. It does not use this in a
BCI, but suggests that it is possible.
It is required for week 4.
Parra et al (2003): This is the
first and only paper in which ERNs
were used to improve BCI performance.
It is required for week 4.
Kubler et al. (2002): This paper
shows that TMS (transcranial magnetic
stimulation) can influence SCPs. It is
recommended for week 4.
Week 5 readings: The week 5 readings
are the slideshows, below, from the
two guest lectures on Tuesday, July
13. You are also required to see
Spiderman 2, since the class voted
that this would be the BCI movie of
interest.
Week 6 readings:
Kennedy et al (2000): In this
paper, locked in patients used an
implanted BCI to control a virtual
keyboard. It is required for week 6.
Wessberg et al (2000): This shows
monkeys controlling a robot arm in 1D
and 3D. It is required for week 6.
Taylor et al (2002): This paper
compares performance of an implanted
BCI used to drive a cursor vs. one to
drive a robot arm. It is required for
week 6.
Kennedy et al (2004): This shows a
successful BCI driven by local field
potentials (LFPs). It is required for
week 6.
Leuthardt et al (2004): This is
another recent paper showing a
successful BCI driven by local field
potentials (LFPs). It is required for
week 6.
Carmena et al (2003): This paper
shows an implanted BCI that enables
monkeys to make both reaching and
grasping movements. It is recommended
for week 6.
Keirn and Aunon (1990): This is
required for week 6. It was the first
paper to explore different mental
tasks that could be used in BCIs and
suggest that they may be useful for a
communication system.
Millan et al (2004): In this
paper, two human subjects moved a
robot around a maze using a mental
task BCI. This is required for week 6.
Penny et al (2000): This paper
showed an online MT BCI for cursor
control. It is required for week 6.
Weiskopf et al (2004): This paper
suggests that an online fMRI BCI based
on mental tasks could be feasible. You
do have to know this fact.
Otherwise, this is recommended reading
for week 6.
Also...
BCI movie vote: Prepare to vote
for the BCI themed movie to be shown
during the "BCI and Culture" lecture
on July 15! All enrolled students get
2 votes. Students who are sitting in
may not vote, but are free to lobby
those who can vote.
Practice midterm questions: Some
of these cover lecture material not
yet presented.
Sample short paper: This is an
example of a short paper in response
to one of the required readings.
Sample project presentation: This
is a presentation from a student in
the BCI class last year. It is
presented here with his permission.
Midterm with answer key: This also
includes some comments on how people
did on the questions and a breakdown
of the scores.
Post midterm comments: Various
comments on grades, the plan for next
week, and the term project.
Last year's midterm: The complete
and unedited midterm from last year.
Practice final exam questions:
Includes comments on the final and
last year's midterm.
Final exam with answer key: As
promised, it is up just after the
final exam. *** The last page of this
key includes the final exam grades as
well as the midterm grades. Since some
of the projects are not yet turned in,
I am not going to put project grades
up yet. Final exams may be picked up
at the GSU BrainLab***
-Brendan 7-30-04 10:00 AM
Intro to BCIs: This was the
slideshow used during the first class
lecture. It provides an overview of
BCIs, why they are needed, some
cultural references to BCIs, BCI
system architecture from the Wolpaw et
al 2002 article listed above, examples
of P300 and mu BCIs, a quick overview
of other BCIs, comments about
challenges for future BCI designers
and mainstream BCI applications, and
sources of more information about BCIs.
Intro to EEGs: This was one of the
slideshows used during the second
class lecture. It provides an
introduction to neurons, why neural
activity produces detectable
electrical activity, micro and macro
electrodes, a typical EEG session,
EEGs vs. ERPs, some examples of
brainwaves, other imaging techniques,
a brief history of EEGs, and some
brainwave applications.
BCIs vs. lab EEGs: This was shown
in the third class lecture. It
describes different types of noise
typically seen in EEG recordings. It
also discusses some means to reduce
noise in both laboratory and field
recordings.
Class presentation
guidelines: This was first shown
in the second class lecture. It
contains some details about the term
project and class presentation to
present that project.
BCI Ethics: This was shown in the
third class lecture. It presents some
ethical questions for BCI research.
Mu
BCIs: This was shown during the
third and fourth class lectures.
P300 BCIs: This was shown
during the fourth and fifth class
lectures.
SSVER BCIs: This
was shown
during the sixth class
lecture.
SCP
BCIs: This was shown during
the seventh class lecture.
ERNs and BCI applications: This
was shown during the seventh class
lecture.
Kennedy slideshow:
This is the slideshow from the Kennedy
lecture. Dr. Kennedy gave a guest
lecture during the eighth class
lecture. The last slide includes
comments on what you need to know.
GSR
lecture: This is the slideshow
from the guest lecture from Umang Dua,
who gave a talk during the ninth class
lecture. The last slide includes
comments on what you need to know.
Assistive technologies: This is
the slideshow from the guest lecture
from Adriane Davis, who gave a talk
during the ninth class lecture. The
last slide includes comments on what
you need to know.
BCI Terms: This was shown
during the eleventh class lecture.
Invasive BCIs: This
was shown
during the eleventh class lecture.
Mental task BCIs: This was shown
during the twelfth class lecture.
BCI-like systems: This will be
shown during the thirteenth and final
class lecture.
Future directions in BCIs: This will
be shown during the thirteenth class
lecture. It includes a discussion
about BCI applications.
|
