BRAIN Interface
BRAIN Interface
BRAIN Interface
CSE 408N
INTERFACING SESSIONAL
2
2. History:
In 1929, a German doctor named Hans Berger announced his
discovery that it was possible to record the electrical impulses of the brain
and display them graphically on paper. He also discovered that these
electrical impulses changed according to the brain's activity, whether in
sleep, under sedation, with lack of oxygen, and in certain neurological
disorders like epilepsy. His discovery laid the groundwork to the field that
today is known as clinical neurophysiology and the basis of today’s BCI
technologies.
The development has gone a long way since then, though still is in
elementary level. Extensive research in this subject is going on all over the
world. NASA has formed a research group named Extension of Human
Senses. Their current work is focusing on using brain waves to control
computer software and the algorithms to support this work.
3
3. Problem Description:
In our project we will develop a brain computer interface that will
enable us to type in the computer through deeply concentrating in a special
type of Key-Matrix displayed in the computer monitor.
4
5. Signal Flow Diagram:
5
6. Description of Hardware Components:
Electrodes are piece of wire or metal pins that works like an antenna to
capture the brain’s electromagnetic waves.
6
6.3. Circuit Embedded in Electrode:
100
+
3V
-
Electrode plates
VGND
TLC272 10nF
R_decupling
200k 2 +5 100
1
Output
3 6 Buffered BW Point A
100pF
+
100pF 2V
-
10k
Bootstrapped HF rejection
Step 1: First of all, boards are etched because industrial etching is expensive
and requires a careful final design. .
Step 2: Pattern is printed using the darkest laser printer settings, onto a sheet
of polyester paper.
7
Step 3: Pattern is cut out, leaving a bit of extra paper, all the way around the
pattern.
Step 4: Copper plate are scrubbed with an abrasive pad. Pattern are laid flat,
face-down, onto it.
8
Step 10: Then holes are drilled with a drill machine using very narrow drill
bit.
Step 13: Chips are placed on the board and then soldered.
Step 15: Then some double cables with shielding were used. One of them
carry the current to the OpAmp (wire +, shielding -), and the other the output
signal. Wires are connected according to the figure.
9
6.5. Electrode Placement:
10
6.6. The Main Amplifier :
Specifications:
High Input Impedance
High noise rejection
Amplify 10-100μV
input signal
Minimal power
consumption
Low cost
11
6.8. Active Low Pass Filter:
CMR LPF
Signal (Noise Filter) (BW Band Filter) m(t)
12
6.9. Circuit Diagram of Main Amplifier & Low pass Filter:
+V
IN4148
.1uF
Point A
+v
Electrode 1 1k 277
1k 22uF
Input 2+ 1 5 +8 7
7
6
3 8
+ - Output
6
4
INA114 1k 4 Point B
10k 100 277 2 1Mega
-V
1 100pF 10k
+ .1uF
3
1k
Input
.001uF
-V
Point A
Electrode 2 10k
IN4148
Now we have the amplified and filtered EEG signal captured from the
electrodes. Now we need to convert this analog brain wave signals to a
digital form to interface with the computer. But the conversion of EEG
signals to digital form would have introduced a more complex circuit for the
EEG. So for simplicity and lower cost we have used sound card to do the job
here.
13
6.10. Hardware Modulator:
14
6.11. AF Band Modulation circuit:
+V
200 .1uF
27k 27k
U5
2
MC1496 3
8 6
10 12 2N2222 22uF
Input 1 5
4 14
+ - T1104
Point B
20k
100 Mic
39k .1uF
39k
SoundCard Mic-In
-V
-V
15k
LM2931-5
1 8 150k
+V VIN OUT 2 NE555 7
3 6
GND 4 5 50k 40%
.1uF
+ + 22uF 12k
22nF .1uF
- 22nF .1uF
- +
-
.1uF
-V 220
- .1nF
22uF
+
1000pF
15
+V
10k
+
9V
2N2222
12k
C22
22uF 22uF .1uF
-V
16
envelope is the original brain wave and that wave is then analyzed further to
decide the mental processing going on the users mind.
6.13.P300 Detection :
P300 is a positive wave that occurs about 300 milliseconds after the
onset of the target/rare (oddball) stimulus among a sequence.
Frequent
Properties of the spike
Rare
is:
polarity (positive)
latency (>275ms)
morphology
(identifiable peak)
17
6.14.Features of P300 :
Central region, about 300ms latency
12
n o n-targ e t s tim uli
ta rg et s tim u li
10
6
A m p litu d e ( u V )
-2
-4
0 10 0 20 0 300 400 500 60 0
(a ). A v erage d w a ve o n C z
18
Flashing of rows/columns containing the desired letter will elicit P300
response at vertex
19
7.Software Description:
The software part starts from the sound card signal recording. Here we
have used some audio libraries for Microsoft visual C++ 6.0 to record the
sound card microphone input signals. After recording the input modulated
signal the sampled data is saved in a temporary buffer. Then from the buffer
our program starts drawing a frame in the OpenGL graphical environment
using the GLUT library. The frame contains the plotted graph of the EEG
signal received, a waterfall model of the amplitude spectrums and also does
necessary calculations to produce the envelope. Then the program also
computes the necessary parameters to detect the P300 spike.
In the mean time there exists a Master Controller program that sends
necessary commands to another PC using UDP protocol to control the user
display feedback. In that computer running the display feedback program
listens to the Master Controller and flashes the display rows/columns
accordingly. The Master controller knows which row/column is flashing in
that PC now. Now the P300 detection subsystem sends information to the
Master Controller if a P300 spike is detected. So Master Controller knows
which row/column was flashing 300ms before the spike was detected. So
after the row and column is detected we can find out what was the
Letter/Number the user concentrated. Now the Master Controller sends
commands to the third PC by using UDP protocol to redirect the
corresponding “Key” for the letter/number detected.
18
9. Brain Computer Interfacing Applications:
69
10. Future Extensions & Fictions:
Cost of parts and equipments for our project is Very high. Again, no
financial support from authority. We have designed a BCI System where
the hardware part will be made of spare home appliances, cheaper
alternatives of the parts/equipments.
Unavailability of some equipments and parts in local markets also create
problem to us. We were delay to start our project for this problem.
However our project supervisor helped us to buy these parts & chips
from Thailand.
The conversion of EEG signals to digital form would have introduced a
more complex circuit for the EEG. So for simplicity and lower cost we
have used sound card to do the job here. The interfacing part of the
digitized signal to the PC is also simplified by this.
Most audio systems, including sound cards, filter out frequencies below
20 Hz. The signals we want to capture are mostly below this frequency,
70
so we can not just plug in the EEG amps and go. To solve this problem
we at first modulated the signal and then pass it through the audio port of
sound card.
Since we used cheaper alternatives of electrodes their performances were
not good. So, expected results were not found from them.
13.References:
71
John Williamson3, Roderick Murray-Smith3,4, Klaus-Robert M¨uller1,2
1Fraunhofer FIRST (IDA), Berlin, Germany
2Technical University Berlin, Berlin, Germany
3University of Glasgow, Glasgow, Scotland
4Hamilton Institute, NUI Maynooth, Ireland
E-mail: [email protected]
72
[13] The OpenEEG Project
http://openeeg.sourceforge.net/doc/
73