Links for current BCI projects

http://bci.epfl.ch/
http://ida.first.fhg.de/projects/bci/bbci_official/index_en.html
http://www.engadget.com/2006/03/08/the-brain-computer-interface/

A new start..

After quite long time, it's been ages since I last posted a thread here. In general, each thread must have a reason; all previous threads were a contineous threads update, but then the project was accomplished. It felt good again on day to write things over, but again; it's one day only.

Now, am back with a new project of interest. Shall I say couple of projects or one project, I still don't know; I can just say, that's a good start for the rest of my life.

Project details will be out soon.. I need to work now; but updates are coming up soon.

_ t a r e K

How it feels reading the whole blog once I am done.

It feels quite amazing reading everything. It's been almost 10 month since I first wrote the first entry in this.

While doing my project, I went through a lot of ups and downs that made me want to stop the project, I even got comments from the professor that I put myself into troubles that it's almost impossible to go out of it! God how I loved that day. It was the day that proved to me, I will either be able to do something, or nothing at all. And I started working.

I am thankful to God that made me be able to go through my project through my ups and downs.

Thankful to Dr. Assaf, of whom without, I don't know what would have happened.
Thankful to Dr. Auda, supervisor of the project and inspirer.
Thankful to all people that believed in me before my project even started and while my project. Thanks Saty!
Thankful to my friends for their support.
Thankful to Ala'a and Mahmoud for their help in data acquistion.

I look back at how the past four month of my life were, they were one of the best times of my life. Not because of my graduation project, but because it was the last semester for me at AUS

I can't write anymore. I am listening to 5edny layk by Wa'el Kffour, I think this is how it's gonna be the ending of my blog. I miss you AUS...


Tareq Mandou
----------------

Presentation Poster 

wavelets! 

Samples of signals 

And finally, MatLab rules!!!! 

Project logo by Nawras (http:\\nawras.deviantart.com) 

Another angle 

How the artificail arm 

Subject After 

Subject before... 

Our subject undertaking the process of signal-acquisition 

Our subject undergoing the process of reading brain-signals. 

How it feels after the presentation

God, I am finally done with my presentation. One of those little feelings that make you feel acomplished in life. Omnia gave me her laptop, Baseel was with me all the time, eman asked about how I did once I finished straight away. Those are the few real people that were there. Abeer, Diyar and Sarah are another good ones. When I went out, all the people congratulated me, that felt amazing. :D

People from different places are asking me how did it go, Som3a called me as well, Zaki did as well.... I have a lot to say but I think I will do some when I get back to my room...

Now, it's 2:54 AM, I feel like writing. I don't know how good or bad my presentation was, but one thing I am sure of is that I tried all I can.

I think I was able to accomplish what I wanted of delivering the idea to professors. After I went out, the computer science gang had launch all together.

I don't know how good or bad I feel, although most of my friends were not there; though it seems they were busy with their scheudel. Baseeeeel came, Nadeem, Wissam Misto, Dano, Jojo & their friend, Zack couldn't make it :(, Aman couldn't due to timing error, Yara n Fares couldn't either, most of UAE girls!!

Winning the NAC compition

El 7amdollelah, today I won the compition, I am happy. I sat a goal to myself and did actually achieve it :)

It feels the best, I am sorry for the rest. :(

From Dr. Jalal's website...

"The most important motive for work in the school and
in life is the pleasure in work, pleasure in result, and the
knowledge of the value of the result to the community"

Albert Einstein

A letter to a teacher

I would like to thank...

Dr. Yousef Al-Assaf: for going all the way through in the Signal Processing phase
(Signal Processing)

Dr. Khaled Assaleh: for confirming that my data is not corrupted
(Signal Processing)

Dr. Tamer Shanableh: for motivating me to get higher results
(Signal Processing)

Dr. Gasser Auda: for origination such an innovative idea and supervising the work
(Neural Network)

Hania (B.S.C.E) for her contribution in the database signals

I believe that each one of these people has contributed not only partially but also in the process of getting these results. Each one of them inspired me in his own way, methodolgy and/or way of thinking.

Many thanks for youe help. Your help is not without gratitude.

Tareq,

New BCI Group...

http://groups-beta.google.com/group/dimensions-bci

Officially... 92.5%

Finally, 92.5% recognition.... Project is done.... :D

After long time...

I can't even remember what I have last written about... It's been quite sometime and at some point, I felt that my project worths nothing at all. After talking to Dr. Tamer abot what I reached to, I felt that I should not expect high expectations. To be honest, I was disappointed at some point. Specially that he told me that it might be difficult for Dr. Khaled to help me out after this long time.

At first, I was disaapointed, but I felt that I started working even harder, although my other courses are being effected, though I decided to finish it with the highest expectations possible.

I am currently on the neural network part, the best part about neural network is that it is all about programming, something that I adore. I did not get very good results at the beginning, though things started getting better, The highest percentage I got so far for the training set while training each class into two or three subclasses is 67.76% Now, I am going to do the approach Dr. Khaled taught me. He said I should be consistent with the data distribution. I actually thought of this, but it would be nice to have your ideas confirmed
I hope this would work well.

Approach used in Voice recognition

After giving up and deciding on not working on the project for sometime, I saw Akram today. He showed big interest in working in the project and the output of the project. He was so much interested that actually made me interested in working in it all over again... The first advice was to talk to Dr. Assaleh, he is so much busy all the time, but I will try to manage and talk to him, inshallah.

Anyways, he told me about an approach that he has used in voice recognition. I liked the approach cause it seemed to make so much sense to me, especially that I have not thought of it that deep.

The approach goes as follow

1. Divide your raw signal into four partiontions
2. Apply FFT to each of these partionions
3. Divide their four FFT data into 8 other samples.
4. Find the energy of each of those 8 FFT samples and thus for each signal, you will have 8 energies for each part --> 4 * 8 = 32 different numbers representing one signal...

When you get a new signal, find the distance of the data and check the closest.

features extraction...

Featuers Extraction:

• FFT, Energy, LPC, Data Distribution, ICA, Classify, AR, Artificat
• Find variablility for each of them

Work on the relation found the following in each of the methods

1. CommonPeakValue: common among all signals
   
2. NumOfPeaks: ranges
   
3. Amplitude, latency,
   
4. Check sliding widnow approach on FFT, LPC, Energy(I doubt it would work on this), AR, Artifcat,
   

Parts concerned with the process

• Gathering Data: sampling technique
• Signals Processing: depends on how good sampling technique is
• Neural Network: Extracted features from signal processing

Signal Processing:

Feature Extraction: ICA, FFT, AR

Classification: amplitude, topography, latency and frequency

Once done, feed all data into Neural Network

Refinement for Final Project Plan... # 3

This approach actually addresses the main problem of the project. Since the problem falls in the following two categories

1. Ranges/amplitude and different signal characteristics of the same action are not similar and/or do not seem to have anything in common

· We don't know what a signal means by normal testing.

2. Number of signals available is small and we have only 1 channel record.

· Cannot generalize an algorithm that would work for all of the signals

In order to over come this problem. I created a neural network for each signal set. Called XXXSignalRange where XXX = EXT, RET, PRO, SUP

To understand further, I'll clarify by an example;

EXTSignalRange

Testing phase:

Takes as a correct input all 20 EXT readings

Takes as an incorrect input all 60 non-EXT.

Training phase:

Rejects three averages as incorrect

Accepts EXT average as correct

------------------------------------

This will happen for each signal.

Thus we would have a network of neural network that predicts what each signal's features would be and what to expect and what not to…

This would give us a better classification so that we can later work with the biggest neural network.

ELE approach

Yesterday, I sat and talked to a friend in Electrical Engineering. I actually got convinced after all that the only way to solve my problem would actually by through Signal Processing... ma 3aleena...

First, I helped my friend with neural network before she helped me with Signal Processing. She taught me what Fast Fourier Transform, LPC and Wavelet represt. Knowledge is amazing... I finally knew what signal processing is about. Make more sense of signals that have nothing to do with time vs. amplitude, but rather frequency vs. amplitude.

That is not the only thing, the better part is actually using the code I have written before to do the same thing.

Dr. Kassem has always said: "You don't have to reinvent the wheel" :)

Anyways, I am going to the lab to work on more effecient processing. I also started working on Neural Network all night yesterday.

Yesterday I learned:

• What signal processing is about?
• Why we need it?
• How to write a Neural Network code. ("reusing what is available")

I also learned that there is always 10,000 ways for the project not work before you reach to the optimum solution...
"I have not failed, I have only found 10,000 ways that won't work" Edison!

Future BCI...

"Your honor, I didn't punch him, my neural implants misinterprieted my intentions. I was actually trying to scratch my butt."

Final project design...

Yesterday, while in Security class. A lot of new ideas came to my mind... Now that I am faced with a major problem of having a similar signal for the same action for the same person does not exist.

• I will create a neural network that would predict the behavior of signals of a specific action so that these signals would be entered into the bigger database.
• Each movement will have a neural network that would predict its behavior and would exclude other signals (e.g. signals for the other three movements.)
• Now that we have 4 trained neural networks, we can work on creating the biggest multilayered neural network that have all set of data and can classify signals into one of the four different actions.
• In case a sample had two similar results, it would solve it according to other methods... (time averaging...)
  
• I will also make the neural network adaptive in the sense that it learns more with time (check self-learning neural network (Dr. Barlas notes!) )
  

Testing a new method...

Today I was working on my Security assignment. It says, if you are a computer scientist, you do the homework the night before it's due. :) After hours, I went to mecatronics and currently finished working on the enhanced sliding window approach. I need to test it once I get back... I am not assuming that it give me the best results, but I am positive that along with the 2nd method "Averaging Signals" it would give a good combination for solving such problems.

Quote of the day
"There is always a mathematical approach to all problems"

Methods I am going to use...

I planned on doing the following today...

Things I need to change...

Come up with at least two new different approaches each week.
Full documentation on these approaches.
Try to come up with approaches that would give a definit result, not an estimated value...

Work on enhancing the different four different methods and approaches created so far.

Redefining Project Objective,
Work on fininding different methods for classifying a movement.

If there is any error with the signal, then there must be no errors with the data brought from the BCI compition.

It's 1:36 am and i am working in Mecatronics lab...

Time to work...

My project is a disaster

Today, I felt lost. My project is a disaster. I am way behind, I don't even have enough data to start working on my project, even if I do, data might not be effecient. So I planned on doing the following...

• Work on fininding a mathematical approach for this problem

Work on other project, work on classifying data that came from other big laboratories in the world and work on processing them.

Today I found out that in order to have correct data, or at least something that we can work with in order to compare. We need to have the 64 channels recordings.

I will talk on Saturday inshallah with Dr. Auda, Dr. Assaf and Dr. Assaleh in order to be able to scopr my research further.

My project is a disaster

Today, I felt lost. My project is a disaster. I am way behind, I don't even have enough data to start working on my project, even if I do, data might not be effecient. So I planned on doing the following...

• Work on fininding a mathematical approach for this problem

Work on other project, work on classifying data that came from other big laboratories in the world and work on processing them.

Today I found out that in order to have correct data, or at least something that we can work with in order to compare. We need to have the 64 channels recordings.

I will talk on Saturday inshallah with Dr. Auda, Dr. Assaf and Dr. Assaleh in order to be able to scopr my research further.

New Ideas...

• Try your algorithm in BCI datasets.
• Classify your work into Training Database and Learning Database.

Three different processes.

Processes to be taken into consideration

Signal Processing
Sampling Processing
Neural Network and Adaptibility...

**New Methodology**

A thought just came to me. Ask the person to do a movement and we will tell the person whether the movement he did was meant to be what the program output gave us.

If yes, add to the DB, else, know what is the meant signal and add it to the desired DB...

This will help in getting right results.

Presentation and Paper Plan

• Work on finding a relationship in the data provided
  

• Explain the effeciency of this relationship.
  

• Explain how abstract this relation could be and prove that it does not work with different signals from different subject.
  

• Explain some of the drawbacksk of such implementation.
  

• Different brain states
• At which state could the signals be read
  

• Define all the problems and how they could be solved
• Suggest solutions
• Future work
  

Albert Einstein

• "Any intelligent fool can make things bigger, more complex, and more violent. It takes a touch of genius -- and a lot of courage -- to move in the opposite direction."
• "Imagination is more important than knowledge."
• "We can't solve problems by using the same kind of thinking we used when we created them."
• "Equations are more important to me, because politics is for the present, but an equation is something for eternity."
• "The secret to creativity is knowing how to hide your sources."
• "If the facts don't fit the theory, change the facts.
• 
  

Third day in lab... I learned.

Today, I wake up late to all my classes and then went to the lab straight away so that I can get some work done. I got tired quite quickly and decided on going back...

Anyways, when I came back at night, it was Nadeem, Bana, Wissam, Wissam and Kinan sitting there and I heard that Hiba and Hadi were there as well. Anyways,as I started working on my project. I started getting some work done. I started trying different ways of plotting.

Food came and I have to run, anyways, I was able to somehow using the function semilogx find some differences, I still need to study it further more. I print it the function details and I am going to write about it once I get back....

I learned a lot, I wil mention once I get back.

Final Project Plan

Today, I went to talked to Dr. Assaf about the new ideas for the project and have finally came to a result about what we are going to do...

• Work on one sample of the data available and come up with a solution
  

• Show that the solution is not optimal and cannot work in some different situations

• Reasons why it does not work with way (e.g. different people different signals, beta-alpha brain states, etc...)

• Summerize problems you faced and proposed solutions
  

• Recommendations
  

• Future work

Now that I know what I am done with, I will start workin on the first phase. The first phase needs learning about different signal processing techniques.

I learned today:
Not to interfear with others business and not to come on behalf of the university.
Be reasonable in my dreams with time limit, rather than limiting the dream forever.
Always follow with your advisor, he is your judge after all :)
Talking to people inspires me and two of these people are Dr. Gasser and Dr. Assaf.

Time to get to work.

Second day in lab... I learned.

This has been open for almost two horus, it's 09:04 pm right now.

Today, as I was reading through the data, I tried to find a relation between them and actually checked using template theory how close are two different readings. I got 1/4000 * 100 % does that count. I also wrote the compare_range file which tries using different range but did not work as yet.

I have learned today:

• How come I am trying to work on finding a relation if it does not exist at the first place. I should try a better way for sampling data or getting them at the first place.
• When comparing between two different singals, don't forget the ranges in the x and y axis.

I should:

• Talk to Dr. Assaf about adaptive algorithm development. Ask the questions that I have in mind.
• Ask about the new technique of reading blood pressure while asking the person to move his arm.

I also learned the following functions:

• isequal,
  
• low-pass filtering,
  
• advanced array manipulation,
  
• polynomial functions representations,
  
• and illustrating different functions

Function of different brainwave

The next brainwave category in order of frequency is alpha. Where beta represented arousal, alpha represents non-arousal. Alpha brainwaves are slower, and higher in amplitude. Their frequency ranges from 9 to 14 cycles per second. A person who has completed a task and sits down to rest is often in an alpha state. A person who takes time out to reflect or meditate is usually in an alpha state. A person who takes a break from a conference and walks in the garden is often in an alpha state.

The next state, theta brainwaves, are typically of even greater amplitude and slower frequency. This frequency range is normally between 5 and 8 cycles a second. A person who has taken time off from a task and begins to daydream is often in a theta brainwave state. A person who is driving on a freeway, and discovers that they can't recall the last five miles, is often in a theta state--induced by the process of freeway driving. The repetitious nature of that form of driving compared to a country road would differentiate a theta state and a beta state in order to perform the driving task safely.

Individuals who do a lot of freeway driving often get good ideas during those periods when they are in theta. Individuals who run outdoors often are in the state of mental relaxation that is slower than alpha and when in theta, they are prone to a flow of ideas. This can also occur in the shower or tub or even while shaving or brushing your hair. It is a state where tasks become so automatic that you can mentally disengage from them. The ideation that can take place during the theta state is often free flow and occurs without censorship or guilt. It is typically a very positive mental state.

The final brainwave state is delta. Here the brainwaves are of the greatest amplitude and slowest frequency. They typically center around a range of 1.5 to 4 cycles per second. They never go down to zero because that would mean that you were brain dead. But, deep dreamless sleep would take you down to the lowest frequency. Typically, 2 to 3 cycles a second.

When we go to bed and read for a few minutes before attempting sleep, we are likely to be in low beta. When we put the book down, turn off the lights and close our eyes, our brainwaves will descend from beta, to alpha, to theta and finally, when we fall asleep, to delta.

SOURCE: http://brain.web-us.com/brainwavesfunction.htm

New approach ~ Blood

New approach is checking the blood movement through the arm while in test...

Take a reading at the beginning before anything happens and check the reading.

Let the person think of moving an object that he/she can see and move up it upward. If there is a difference in the blood movement, that is right.
Make the timing for the movement something that is feasible... Get different heights and different masses in order to provide the timing we want.

To make readings even better, try to compare the blood movements between the different trials.

New approach ~ Hypnosis

Yesterday while I was explaining about the concept of my graduation project with the three Junior Engineers, a new idea came to me regarding reading brainwaves.

If we had the subject in a clear state where he thinks of nothing but what we are going to tell him/her, we will be getting the best results. This can be best acheieved using Hypnosis.

Hypnosis will give us the best results for two reasons.

First, it will provide a clear signal since the subject is only thinking of the commands given.
Second, in hypnosis, a person is actually using his beta state of mind which gives us the best results.


Still a lot of reading and research to go... I will have alternatives for sure...

New idea*

The idea of having different nnet is actually that each of them support different weights capacilities and layers.

The idea of that is actually that a set of input of "right signals" are given to the nnet in order to learn what is the range of the inputs and tries to assign different wieghts accordingly so that it will try to give the some output for these data with a slight percentage of error recovery.

That is awesome! The idea is that how u assign these different weights.

About the project

Using hypnothesis to get data

The difference between networks is in using the wights, the new one is the mostly used nowadays...

Training Empirical and theoritical might be different. start working on figuring out whether weight makes the difference

Learn or get someone to make hypnothesis

Project Proposal:
Figure out the possiblity of reading brain signals. Give two different approaches, two different people and that training is the matter
At the end, present the new two solutions with the hypnothesis

Start of the semester

After a week from the beginning of the semester, this is a plan of what I will do

1. Think of a method to do gather data.
2. Meet with students and finalize issue regarding timing and other stuff.
3. Meet with students to sign contract
4. Start taking data.
5. Design the poster. = Malek's help
6. Hania's session for signal taking.
7. My sessions for signal taking once I am done with the procedure.
8. Write a well-formatted-and-documented method for taking the signal... Software for doing so will be a plus
Good luck with rest...

I am meeting up with students and partner tomorrow in order to finalize the timing issue.

What I will do through the semester

While reading a book about Writing and Presenting Scientific Papers. I started thinking of how I used to look at problems and solve them. I miss being the old Tarek, calm one, not so hyper; I should get back to that one....

Anyways, I will write down what I have decided on doing for this coming semester.

• I want to start learning like the old days, I am doing this project because there are people out there that my research will actually help.
• I want learn how things are done, just like the old days, when I used to do in MIS 201, CMP 433, CMP 415, I did a really good job back then. I still can.
• There are set of things that I want to see by the end of this project;

1. To write a fully scientific paper that is as interesting as other writings.

2. To learn what I am doing. I am not memorzing, I want to know how to do it over in a better way

• Write down what I am learning from the book about writing a paper, it also helps me in doing the research as well
• I want to google "How to Write"
• I want Dr. Cindy to read my research at the end.
• Buy a booklet or use this link to write what I want to do and did.

Back home...

Now, I started working on my project. I am currently working on what I have learned in Software Engineering. I want my project to be the best.

I feel that I can do it and I am going the exact steps that I have written ahead of time. I am going to write everything in paper tomorrow so that I can be as creative as possible and I will afterward write down everything I am going to do...

The 4 months phases, and project outline and deadlines... Everything in MS Project...

Tareq

Final Project Phases

I finally decided on what I am going to do with my final project.

First phase:
A normal neural network project for moving the artificial arm! Could be real-time project. This will make it better. This should not take more than a month.

Second phase:
Doing my actual job. Developing my own algorithm... Something I've always liked.

How to write and present scientific papers.

Things are different. I've been reading a book on how to write and present a scientific paper. It is really useful as it mentions not only what you need to write the paper for your graduation, but also how to present it and make these two things go together. I will write furthermore about this book later...

Procedure for my graduation project

My project will be actually to re-define the method in which neural networks are trained. In this case, I will try to do the following:

1. Try to once you have all the data to analyze them through normal Neural Network Training Algorithm

2. Try to make your own result and use your own algorithm in order to analyze the data

3. Compare between the results in both cases.

Meeting with the partner...

I finally got to meet with Hania and Dr. Assaf. We discussed what my project will be about and I am actually so much happy that I got the chance to talk to them. I hope I will be of some help to both of them. Hania has done exactly what I was looking forward on doing. It was amazing in the sense that I could find myself doing my final project.
I have to admit though that it is not going to be easy, but it will require a lot of work. Mos importantly, I am happy now and I hope my project will accomplish what it is meant to to do...

We talked about the arm, the software and she will inshallah provide me with the data...

Time to get to bed...

Topics to be covered while doing the project...

Following is a list of topics that need to be covered while doing the project:
For big amount of data, it will be processed, Neural Netowrk will train the data and once all of that is done, it will be integrated with the arm design which will do the simulation...

Digital Signal Processing

Neural Netowrk

Arm System Controller

Arm Design

All these will be controlled using Matlab Simulink once each is understood seperately and then fully integrated.

Phases continued...

Following is the phases that should be done...

Data Acquasition:
This should be done as soon as possible in order to be able to process data and start training the neural network.

The faster the neural network in being trained the better results we will get and thus the project will not be pure signal processing but will also include a major part of enhancing the neural network itself.

Once Singla Processing is understood, I should find a new way of training the neural network other than the normal way of re-training, or so called, reinventing the wheel...

Processing Signals:
Nothing in mind right now...

MatLab.... Finally

I finally installed MatLab. God, this software is amazing, now one of the goals I've set to myself it to work for MathWorks.com
It has really good stuff that I started thinking of changing the topic I am going to work on. I passed by Dr. Assaf and Dr. Auda couple of days ago, and they are actually thinking of doing something together. The project seems to be tough, I just hope it will be fine.
We went today and discussed the possibility of changing the hardware to a software the processes data. That is going to be amazing. The software name is gBSanalyze.

Good luck!

Getting my own Desktop

I finally got my own desktop, it feels so good, I got it with Ala'a and we assembled it on owr own. Well, on his own :)
Anyways, I started working on it and Media Edition is really good. :P

It's time to get to bed

Final Project Phases

The project consist of 5 different phases that need to be accomplished according to the timeline below.

PHASE I:

Assemble the Robotic Arm Trainer and control it through the PC Interface. The software can be downloaded from http://robotikitsdirect.com/pages/downloads.html. Should make sure that the software and the arm moves very well. Assembling should not take a lot of time.

This phase should be finished by 20-11-2004

PHASE II:

Have a logo for the project. Ask a student from design management or a close friend to desing a logo for this, the logo should resemble Cutting-Edge Tech., Arm, Paralyzed people... Ask Rami, Kareem, Samia or even Melissa to design that.

This phase should be finished before 20-01-2005

PHASE III:

Define the method and/or software in whcih the Arm will be controlled. This can be done in three ways,

1. Using Matlab
2. Using Visual Basic .NET
3. Using the interface provided with the arm.

I am still researching on if it is possible to control it using MatLab without the need for the interface. However, another way can be done though VB .NET which will run the Matlab code, and then control the Arm.

If it was done through (3), the data can be output by the matlab as a text file and then will be loaded into the provided interface so that it can be executed. This is the worst case scenario

This phase should be finished by 15-12-2004

PHASE IV:

Start doing your project.

This phase should be finished by Not now at all...

PHASE V:

Start working on the paper.

This phase should be finished by Not now neither...

Consolidating Data...

There are 2 different systems in which the project might be done:

1. For one sample, have different trails for each action, and once you have the result, compare it with the original one that you have in your Database.

Analyzing New Data:


New Subjects:
2 distinc trials for doing the 5 different actions of moving (right, left, up, down and grasping), named respectively, Trial_test, Trial_serious,
then
1 trial for doing the 5 different actions in the following order (grasping, moving down, up, left and right), named Trial_opposite
it could be 1 - 1 -1, which i believe is better.
last trial would ask the person to think of a random action and confirm the program's result, named Trial_final.

Saving Data:
The readings Trial_serious, Trial_opposite and Trial_test will be added to a Temporary_DB which will be confiremd or disconfirmed by Trial_final.
These trials are then added to the NN_DB for each action respectively.


Old Subjects:
Compare the signal Trial_signal with the different action signals avg_signal_right, avg_signal_left, avg_singal_up, avg_signal_down and avg_signal_grasp.
If there is a result with less than 10% difference between two different avg_signals, the Trail_signal will be compared with the different signals of each of the different actions.

Welcome to Brain-Computer Interface

This will talk about all useful Brainwave Computer Interface aka "BCI" that I can get from the net or from reading books. All references can be found in the list of links. This blog is part of my graduation project.
This blog will contain, hopefully by the end of the term, everything related to my graduation project from A to Z. Today is 07-11-2004. The graduation project should be next semester, however, I started now as the advice I took from a friend in order to finish the project on time with the maximum requirements I can make. I will post later my scheduel and timeline. I hope that I'll be able to stick to it...