Brain
What is a BCI?
Brain computer Interface is a system that inputs or outputs information from brain. It captures and transforms signals originating from human brain into commands that can control devices and applications. BCIs modulate signals in human brain to add information or change mental state or behavior.
What is an EEG?
Electroencephalography (EEG) is the recording of electrical activity along the scalp. It measures voltage fluctuations resulting from ionic current flows within the neurons of the brain.
What are the physical principles of fNIRS?
➢ Tissue is transparent to light within the near Infrared (NIR) range, hence it can penetrate it.
➢Main chromophores (absorbers) within the NIR light range are oxygenated and deoxygenated hemoglobin as compared to other tissue components such as water, lipid, melanin, etc.
➢ Oxy- and deoxy-Hb have distinct absorption characteristics within the NIR range which allows spectroscopic measurements
➢Monitoring oxy- and deoxy-Hb concentration changes indicates ➢Oxygen consumption ➢Blood volume change
What are protocols? List the common BCI protocols.
Protocols are mental tasks required to engage with the BCI. They generate a unique Brain Activity (signal), and can be an indirect (secondary) process.
Common BCI protocols are: VEPs, SCPs, ERPs, MI, Neurofeedback / Direct Control, Cognitive Workload, Mental Arithmetic / Math, Working memory, Verbal Fluency, Mental rotation, Music Imagery, Speech generation/imagery, SMRs, P300 speller, and SSVEP
What is machine learning? How do we apply ML?
Machine learning is teaching a computer about the world.
We apply machine learning by observing the world, developing models that match observations, teaching computer to learn these models, and then the computer applies learned model to the world.
Which lobe is where a person has sensation?
Parietal lobe
What are Excitation and Inhibition?
Excitation - Increased spike rate, depolarization, excitory post-synaptic potentials
Inhibition - Decreased spike rate, hyperpolarization, inhibitory post-synaptic potentials
Photos that enter living tissue undergo what two types of interaction?
- Scattering (cell membranes): causes photos to change their direction of motion due to cell membranes and tissue boundaries. Assumed to be constant
- Absorption (Hb, HbO2, water): causes the photos to lose their energy to the medium due to chromophores in the tissue. Changing depending on the changes in the concentrations of the chromophores.
What are slow cortical potentials?
Slow cortical potentials have the lowest frequency features of the scalp recorded EEG. Negative SCPs are associated with movement and other functions involving cortical activation. Positive SCPs are usually associated with reduced cortical activation.
What are Terminologies 1 & 2?
Terminology 1
• Feature: a variable (predictor) believed to carry discriminating and characterizing information about the objects under consideration
• Feature vector: A collection of d features, ordered in some meaningful way into a d-dimensional column vector, that represents the signature of the object to be identified.
• Feature space: The d-dimensional space in which the feature vectors lie. A d-dimensional vector in a d-dimensional space constitutes a point in that space.
Terminology 2
• Class: The category to which a given object belongs, typically denoted by ω
• Pattern: A collection of features of an object under consideration, along with the correct class information of that object. In classification, a pattern is a pair of variables, { 𝑥 Ԧ , ω} where 𝑥 Ԧ is the feature vector and ω is the corresponding label
• Decision boundary: A boundary in the d-dimensional feature space that separates patterns of different classes from each other
Which lobe lies in its deep structures the hippocampus
Temporal lobe
What are the Paradigms of P300?
Visual
• Row/column
• Single character/cell
• Checkerboard
• Region based
Audio
• Spatial/location of sounds
• Frequency/type of sounds
Time domain (or time resolved) systems for fNIRS?
Input: Extremely short incident pulses of light (picoseconds) Output: Broadened and attenuated version of the incident light (due to absorption and scattering.)
Advantages Spatial resolution, Penetration depth, Most accurate separation of absorption and scattering
Disadvantages: Sampling rate, Instrument size/weight, Cost
What are sensorimotor rhythms?
• In awake people, primary sensory or motor cortical areas often display 8–12 Hz EEG activity when they are not engaged in processing sensory input or producing motor output (mu rhythm)
• They are associated with those cortical areas most directly connected to the brain’s normal motor output channels.
• Movement or preparation for movement is typically accompanied by a decrease in mu and beta rhythms, particularly contralateral to the movement. This decrease has been labeled ‘event-related desynchronization’ or ERD
• Can be asynchronous
• Useful for continuous systems
• Based on changes in band-power
• Event-related synchronization (ERS)
• Event-related desynchronization (ERD)
• Spatial patterns can be used to easily create multiclass BCIs
What are Terminologies 3 & 4?
Terminology 3
• Training Data: Data used during training of a classifier for which the correct labels are a priori known
• Test / Validation Data: Data not used during training, but rather set aside to estimate the true (generalization) performance of a classifier, for which correct labels are also a priori known
• Field Test Data: Unknown data to be classified for which the classifier is ultimately trained. The correct class labels for these data are not known a priori.
Terminology 4
• Cost Function: A quantitative measure that represents the cost of making an error. The classifier is trained to minimize this function.
• Classifier: A parametric or nonparametric model which adjusts its parameters or weights to find the correct decision boundaries through a learning algorithm using a training dataset – such that a cost function is minimized.
• Model: A simplified mathematical / statistical construct that mimics (acts like) the underlying physical phenomenon that generated the original data
Which gyrus lies in the frontal lobe?
precentral gyrus
Name and describe EEG bands
Delta - up to 4 Hz, associated
Theta - 4 to 8 Hz, associated with childhood, drowsiness
Alpha - 8 to 12 Hz, relaxed, alert state of consciousness, present at 2 years old
Beta - 12 to 26 Hz, anxious thinking, active concentration
Gamma - above 30 Hz, perception, problem solving
What are types of fNIRS signal processing?
Motion Artifact Detection & Removal
• Low-pass / Band-pass Filters
• Wavelet Analysis
• Independent Component Analysis (ICA)
• Principle Component Analysis (PCA)
• Coefficient of Variance related (SMAR, etc.)
• Optimal Filtering (Adaptive; Wiener; Kalman)
What is neurofeedback/direct control? And what is cognitive workload?
Neurofeedback/Direct control - using localized brain activation, provided to user in a closed loop system (usually visual feedback), users learn to self-modulate specific brain areas
Cognitive Workload - additional computation/processing that is performed deliberately to signal a command (and absence signals another command)
What is Terminology 5? What are the examples?
• Parametric Model: A probabilistic / statistical model that assumes that the underlying phenomenon follows a specific known probability distribution. The parameters of such a model are the parameters of the distribution.
• A classifier based on determining the parameters of a distribution is also called a generative model as the underlying distribution can be generated from the parameters.
• Examples: Bayes classifier, expectation-maximization algorithm.
• Nonparametric model: A model that does not assume a specific distribution, and that typically follows an optimization algorithm to minimize error.
• A classifier based on using a nonparametric approach is also called a discriminative model, as the decision is then based on a discriminant (or discriminant function).
• Examples: Neural networks, decision trees, support vector machines.
What are all the categories?
• Operation - Active, Reactive, Passive
• Coupling (sensor, electrode) - Invasive, Non-invasive
• Flow of information - Read-only, Write-only, Read-write
• Target - Clinical, Military, Nonclinical/Healthy
• Application - Control, Communication, Prosthetics , Rehabilitation/Therapeutics, Training, Gaming, Art, HCI, Cognitive state monitoring
• Modality - Unimodal, Multimodal
• Timing - Synchronous, Asynchronous/Self-paced
What are the strengths of EEG?
• EEG has very high temporal resolution (typically 2 ms)
• EEG is best suited to hypotheses about time and frequency.
• Speed of processing, Relative order of processes, Temporal relationships (correlation, functional connectivity)
• EEG measures electric potentials
• EEG signals can be used in many ways:
• ERP, Frequency, Time/Frequency
• EEG is best-suited to capture fast brain signals
• EEG can provide spatial information
How does fNIRS compare to fMRI?
fNIRS
• Weaker Contrast-To-Noise Ratio (CNR) than fMRI
• Lower spatial resolution
• Higher temporal resolution
• CNR of fNIRS relative to fMRI dependent on scalp-brain distance
• And, low-cost, silent, repeatable & no contraindications
What are ERP types?
P300
SSVEP
Cognitive Workload
Sensorimotor BCI (EEG)
Motor Imagery BCIs (fNIRS)
What are the supervised and unsupervised algorithms?
Supervised
• Nearest Neighbors
• Decision Trees
• Linear Discriminant Analysis
• Artificial Neural Networks
• Support Vector Machines
Unsupervised
• K‐Means
• Gaussian Mixture Models