User:FabioZennaro

From AIRWiki
Jump to: navigation, search


Fabio Massimo Zennaro
Foto di FabioZennaro
First Name: Fabio Massimo
Last Name: Zennaro
E-Mail: fabio.zennaro@mail.polimi.it
Advisor:
Project page: BCI based on Motor Imagery
Project page(s): BCI based on Motor Imagery
Status: inactive

I'm a fifth year student in Computer Engineering in Politecnico di Milano. I got my undergraduate degree in Computer Engineering from Politecnico di Milano in July 2007. I'm currently working on my final project in order to get my graduate degree.

My final project is about brain-computer interfaces, in particular non-invasive sensorimotor rhythm-based BCI. The touchstone of the project is the Wardsworth BCI interface developed by Wolpaw and his

colleagues.



Description of the final project

As now, the current aim of the final project is to implement a working sensorimotor rhythm-based brain-computer interface and

use it to control a cursor on a screen. To achieve this goal, I am working with my colleagues and my supervisor inside the

university lab; our brain-computer interface is based on the free open-source framework

BCI2000.


Timeline of the final project

July

Week 1 (14th-20th):

Week 2 (21st-27th):

  • I was taught how to mount an EEG
  • I learnt how to use and how to configure BCI2000
  • I started reading papers about BCI research [1] [2]


August

Week 3 (28th-3rd):

  • I prepared with my colleague a brief presentation about the use of r^2 and MEM inside BCI2000
  • I performed further EEG acquistions
  • I read papers about BCI research [3]

Week 4 (4th-10th):

  • Project suspended: holidays

Week 5 (11th-17th):

  • Project suspended: holidays

Week 6 (18th-24th):

  • Project suspended: holidays

Week 7 (25th-31st):

  • I performed further EEG acquisitions
  • I read papers about BCI research [4] [5] [6]


September

Week 8 (1st-7th):

  • I performed further EEG acquisitions
  • I read papers about BCI research [7] [8] [9] [10]

Week 9 (8th-14th)

  • I performed further EEG acquisitions
  • I read papers about BCI research [11] [12] [13] [14]

Week 10 (15th-21st)

  • I downloaded the source code of BCI and got my hands on
  • I read papers about BCI research [15] [16]

Week 11 (22nd-28th)


October

Week 12 (29th - 5th)

  • I tried modifying BCI2000 sources and recompiling it
  • I read papers about BCI research [21] [22] [23]

Week 13 (6th - 12th)

  • I started writing draft chapters for my final project (introduction, brain activity detection techniques)
  • I read papers about BCI research [24] [25] [26] [27]

Week 14 (13th - 19th)

  • I wrote a draft chapter for my final project (performance evaluation techniques, r^2)
  • I started working on raw data inside Matlab
  • I read papers about BCI research [28] [29] [30] [31]

Week 15 (20th - 26th)

  • I wrote a draft chapter for my final project (frequency analysis)
  • I read papers about BCI research [32] [33] [34]


November

Week 16 (27th - 2nd)

  • I studied how the autoregressive spectral estimation works
  • I carried on writing the draft chapter about frequency analysis
  • I read papers about BCI research [35] [36] [37]

Week 17 (3rd - 9th)

  • I converted my notes and my draft chapters to LaTeX format
  • I converted my bibliography to BibTeX format
  • I carried on writing the draft chapter about brain activity detection techniques
  • I read papers about BCI research [38] [39]

Week 18 (10th - 16th)

  • I wrote draft chapters about signal processing and signal classification
  • I read papers about BCI research [40]

Week 19 (17th - 23rd)

  • I read a paper about BCI research [41]

Week 20 (24th - 30thrd)

  • I met a new colleague: Paolo Calloni
  • I discussed with my tutor about the specific subject of my final project
  • I read papers about BCI research [42] [43] [44]


December

Week 21 (1st - 7nd)

  • We met with our supervisors to discuss about our future acquisitions

Week 22 (8th - 14th)

  • We performed EEG acquisitions in order to get initial data to setup BCI applications

Week 23 (15th - 21st)

  • We performed EEG acquisitions in order to get initial data to setup BCI applications

Week 24 (22nd - 28th)

  • I worked on the module development in BCI2000
  • I read snippets of code for feature extraction


January

Week 25 (29th - 4th)

  • I worked on the module development in BCI2000

Week 26 (5th - 11th)

  • We met with our supervisors to discuss about the data we collected during the acquisitions in December
  • We studied the data and signals we collected in order to find out useful features

Week 27 (12th- 18th)

  • We worked on the data and signals we collected

Week 28 (19th - 25th)

  • Project suspended: first-term exams


February

Week 29 (26th - 1st)

  • Project suspended: first-term exams

Week 30 (2nd - 8th)

  • Project suspended: first-term exams

Week 31 (9th - 15th)

  • Project suspended: first-term exams

Week 32 (16th - 22nd)

  • I read about feature selection and feature extraction [45]


March

Week 33 (23rd - 1st)

  • We performed EEG acquisitions
  • I watched lectures online about machine learning and feature analysis [46]

Week 34 (2nd - 8th)

  • We performed EEG acquisitions
  • I watched lectures online about machine learning and feature analysis [46]
  • I wrote code in Matlab for feature analysis

Week 35 (9th - 15th)

  • We performed EEG acquisitions
  • I watched lectures online about machine learning and feature analysis [46]
  • I wrote code in Matlab for feature analysis
  • I worked on classification (binary classifier)

Week 36 (16th - 22nd)

  • Project suspended: I attended a course on 'Collective Intelligence' at Telecom ParisTech

Week 37 (23rd - 29th)

  • I wrote code in Matlab for feature analysis
  • I worked on classification (tree classifier)
  • I worked on feature extraction (Wilcoxon test)


April

Week 38 (30rd - 5th)

  • I wrote code in Matlab for feature analysis
  • I wrote code in C++ for classification

Week 39(6th - 12th)

  • I wrote code in Matlab for feature analysis
  • I worked on feature extraction (SFS, SBS)

Week 40 (13th - 19th)

  • I wrote code in Matlab for feature analysis
  • I worked on feature extraction (genetic algorithms)
  • I performed tests on the performance of feature extraction and classification

Week 41 (20th - 26th)

  • I wrote code in Matlab for feature analysis
  • I worked on feature extraction (genetic algorithms)
  • I wrote a draft chapter for my thesis (state of art)


May

Week 42 (27th - 3rd)

  • I wrote code in Matlab for feature analysis
  • I wrote a draft chapter for my thesis (feature extraction and feature selection)

Week 43 (4th - 10th)

  • I wrote code in Matlab for feature analysis
  • I worked on feature projection (PCA)
  • I started reading about pattern recognition [47]

Week 44 (11th - 17th)

  • I wrote code in Matlab for feature analysis
  • I worked on statistical tests (Friedman, Nemenyi)

Week 45 (18th - 24th)

  • We performed EEG acquisitions with a feedback
  • I wrote code in Matlab for feature analysis
  • I worked on statistical tests (Holm)

Week 46 (25th - 31st)

  • I wrote code in Matlab for feature analysis
  • We performed EEG acquisitions with feedback


June

Week 47 (1st - 7th)

  • We performed EEG acquisitions with the speller

Week 48 (8th - 14th)

  • We performed EEG acquisitions with the speller

Week 49 (15th - 21st)

  • I wrote a draft chapter for my thesis (basics on the brain)

Week 50 (22nd - 28th)

  • Project suspended: second-term exams


July

Week 51 (29th - 5th)

  • Project suspended: second-term exams

Week 52 (6th - 12th)

  • Project suspended: second-term exams

Week 53 (13th - 19th)

  • Project suspended: second-term exams



Bibliography of the final project

Disclaimer: as you can see, this is just a raw draft of the bibliography; I use it only to keep track of the papers I

read, so each entry does not contain all the data related to the source, but only the name of an author (usually the most

famous from each research group, e.g. J.R. Wolpaw for the Wadsworth Center or G. Pfurtscheller for the University of Graz),

the title of the paper and the year of publication.

I highlighted, using a bold font, the most important papers related to my final project.

  • [1] J.R. Wolpaw et alii, Control of a 2D movement signal by a non-invasive BCI in humans, 2004
  • [2] J.R. Wolpaw et alii, BCI for communication and control, 2002
  • [3] T.J. Ulrych et alii, Maximum Entropy Spectral Analysis and Autoregressive Decomposition, 1975
  • [4] J.B. Ochoa, EEG Signal Classification for BCI Application, 2002
  • [5] G. Pfurtscheller, BCI - A brief introductin, 2005
  • [6] J.R. Wolpaw et alii, EEG-based BCI for cursor control, 2004
  • [7] M. Mattioco, Attività corticale e immaginazione motoria, 2003
  • [8] J.R. Wolpaw et alii, BCI Research at the Wadsworth Center, 2000
  • [9] M. Barinaga, Turning thoughts into actions, 1999
  • [10] J.R. Wolpaw et alii, EEG-based BCI communication: system description, 1990
  • [11] J.R. Wolpaw et alii, Multichannel EEG-based BCI, 1994
  • [12] J.J. Vidal, Toward direct brain-computer communication, 1973
  • [13] B. Allison, P300 or not P300: towards a better P300 BCI, 2003
  • [14] J.R. Wolpaw et alii, The Wadsworth BCI Research and Development Program: at home with BCI, 2006
  • [15] G. Schalk, BCI2000: A general purpose BCI system, 2004
  • [16] J.R. Wolpaw et alii, BCI: Detection instead of classification, 2007
  • [17] Hochberg L.R. et alii, Neuronal ensemble control of prosthetic devices by a human with tetraplegia, 2006
  • [18] J.R. Wolpaw et alii, Tracking of the mu rhythm using a derived matched filter, 2005
  • [19] J.R. Wolpaw et alii, Sensorimotor rhythm-based BCI: feature selection by regression improves performances, 2005
  • [20] J.R. Wolpaw et alii, Linear and non-linear methods for BCI, 2003
  • [21] G. Schalk, BCI2000 Manual
  • [22] J.R. Wolpaw et alii, EEG-based BCI for cursor control, 1991
  • [23] J.R. Wolpaw et alii, BCI operation: signal noise during early training sessions, 2004
  • [24] J.R. Wolpaw et alii, BCI Meeting 2005: workshop on BCI signal processing, feature extraction and translation,

2006

  • [25] J.R. Wolpaw et alii, BCI Operation: optimizing information transfer rate, 2003
  • [26] I. Wickelgreen, Tapping the mind, 2003
  • [27] J.R. Wolpaw et alii, EEG-based communication and control: short-term role of feedback, 1998
  • [28] J.R. Wolpaw et alii, BCI Technology: a review of the first international meeting, 2000
  • [29] J.R. Wolpaw et alii, EEG-based communication: improved accuracy by response verification, 1998
  • [30] J.R. Wolpaw et alii, The Wadsworth Center BCI research and development program, 2003
  • [31] J.R. Wolpaw et alii, Evaluation of AR spectral estimation model order for BCI, 2006
  • [32] J.R. Wolpaw et alii, Answering questions with an EEG-based BCI, 1998
  • [33] J.R. Wolpaw et alii, EEG-based communication: analysis of concurrent EMG activity, 1998
  • [34] T.D. Schneider, Information Theory Primer, 2007
  • [35] J.R. Wolpaw et alii, Real-time detection of event-related brain activity, 2008
  • [36] J.R. Wolpaw et alii, Conversion of a EEG activity into cursor movement by a BCI, 2004
  • [37] J.R. Wolpaw et alii, EEG-based communication and control: speed accuracy relationships, 2003
  • [38] G. Pfurtscheller et alii, Current trends in Graz Brain-Computer Interface Research, 2000
  • [39] G. Pfurtscheller et alii, Graz-BCI: State of the Art and Clinical Applications, 2003
  • [40] J.R. Wolpaw et alii, Spatial Filter Selection for EEG-based communication, 1997
  • [41] J.R. Wolpaw et alii, EMG Contamination of EEG: Spectral and Topographical Characteristics, 2003
  • [42] J.R. Wolpaw et alii, EEG-Based Communication: Presence of an Error Potential, 2000
  • [43] U. Hoffmann et alii, Recent advances in BCI, 2007
  • [44] J.R. Wolpaw et alii, EEG-Based Communication: EEG control vs system performance in humans, 2003
  • [45] I. Guyon, Feature Extraction Foundations and Applications, 2006
  • [46] PASCAL Bootcamp in Machine Learning, [1]
  • [47] R.O. Duda, Pattern Classification, 2000