Task 7 Flashcards
Explain the assumptions of fNIRS:
- Active areas need Oxygen
- Oxygen is carried by Hemoglobin
- Hemoglobin absorbs light in a banana shaped way
- Oxygenated HB and deoxygenated HB blood have different optical properties -> They absorb different wavelength
What are the different optical properties of Oxygenated and Deoxgynated blood ?
- > Oxygenated = 850
- > Deoxyenated = 760
- > in general it is the range bewteen 700 -900 nm
Describe the Method of fNIRS:
- > 2 types of wavenlgth inform of near infrared light are introduced by the brain
- > Photons pass trough tissues and are then absorbed by Oxy or Deoxy hemoglobin
- > After absorption the rest of the photons will be refelected
- > The more Oxy is present the less 750 wavelength will be reflected
- > Based on the difference (how much came in and how much came out) we can estimate which regions were active
What is an important rule of fNIRS ?
-> When less Oxy will be reflected then more deoxy will be reflected
What are the limitaions of fNIRS?
-> Does only measure surface brain tissues 2-3 mm
What is an optode ?
= light source of fNIRS
What do you need for an fNIRS scan ?
-> Optode + detector
What happens if we vary the distance between optode and detector ?
- > Higher distance = more depth
- > Higher distance = less quality signal
- > Usually placed 2-7cm away from Optode
- > best is 3 cm
What is meant by event related optical signal ?
-> Measure change in optical properties at cell membranes themselves
-> more a direct meassure (better temporal resolution)
BUt lower SNR
What are more preceise fnirs method and why ?
- Frequency domain systems
- Time-resolves systems
- > both use leaser (more dangerous and expensive larger)
What is a less preceise fnirs method and why ?
- > Continuous wave (CW)
- > Since it uses light instead of leaser
- > Cheaper / smaller / less dangerous
Name some Pro of fNIRS:
- > Two Dependent variables
- > Save / non invasive
- > Low coast
- > Portable
- > Can be integrated with other technologies
- > No vulnerability to electromagnetic environment
Name some Con of fNIRS:
- > Light scattering / absorption by Skin / hair
- > Limited spatial resolution only reaching surface
- > Can not do comparisons of fNIRS data between subject
- > Limitations in the use of cranial reference points
What does fNIRS and fMRI have in common ?
- > Indirect meassurments
- > Both achived similar outcomes
Why is fNIRS worse then fMRI ?
- > Lowers spatial resolution
- > Limited to outer cortex
Why is fnirs bettern then fMRI ?
- > Portable / People can do whatever they want
- > Cheap and quiet
- > Can be intergreatd with other methods
- > Can use metal
- > Ecological validity (real world environment testing)
What is BCI ?
-> System connects a brain with a computer without involving any peripheral nerves and muscles
What can BCI systems do ?
-> Allows for controlling external devices trough brain activity , assisting, replacing cognitive sensory or motor functions
What is the Method of BCI ?
- > Measure brain activity via EEG or fMRI
- > Feature / Signal extraction
- > Signal are translated into a set of rule based algorithm -> Into a machine command
- > Command then gets passed on to the computer
What is a danger to BCI ?
- Low SNR can lead to wrong commands
How do we overcome a low SNR ?
- > Via neurofeedback training -> learn in realt time to change brain activity in desired direction
- > Machine learning algorithms improve accuracy of prediction
Name the two methods which are used as signals
- > Direct encoding (worse)
- > Indirect encoding (better)
- > Both make use of Spatial features/temporal features and magnitude features (certain lvl of brain actvity)
What is meant by direct encoding?
- > brain activity of answering yes or no
- > Bad because = they are heavily correlated therefore not differentiated
What is meant by indirect encoding ?
- > Subsitude Yes resopnse with a mental motor actvity
- > subsitute no response with a inner speach
- > Basically making use of higher order cognitive processes
What is currently the accuracy of BCI ?
-> 70 %
What are some applications of BCI ?
- > locked in syndrome = detecting conscious awareness in non-responsive patient
- > brain signals that an individual can intentionally generate to encode an intention
What is meant by the univariate general linear model ?
- > Single channel of interest is determined
- > Has to calculate a cut of value
What is meant by the multivaraite General linear Model ?
-> Uses all channels
-> Doe not require calculation of cut of value
-> Uses a support vector machine as classifier of a signal
o Which needs to be trained
How can the BCI be used in combination with EEG ?
- > Can use P 300 component
- > Can use Slow cortical potentials & steady state visual-evoked potentials
How can the BCI be used in combination with MEG ?
Real time BCI
- > Can use P 300 component
- > Can use Slow cortical potentials & steady state visual-evoked potentials
How can the BCI be used in combination with fMRI?
- > Automated interpretation & classification of activation maps
- > Neuronfeedback
How can the BCI be used in combination with NIRS ?
-> motor imager or P 300
How to claculate the concentration of Oxygenated and deogynated blood ?
-> via the Lambert beer law
What is one unique benefit of fnirs over fMRI ?
-> Fnirs meassures both oxygenated and deoxygenated concentration while fmri only meassures oxygenated concentration
What is BCI used for ?
- > Repearing
-> supporting
-> Replacing
cognitive function !
Method of BCI 2.0
Signal acquisution (EEG / FMRI) Signal processing (Signal extraction + transforming algorithm)
