Brain Machine Interface - Exam 1

Front 1

Whatis a BMI

Back 1

•Adirect communication pathway between the brain and a machine.m8R4SdN8Bg

Front 2

•Examples of BMI

Back 2

–Neuralcontrol of a robotic arm–Directneural stimulation (feedback) due to touch of the robotic arm/hand–Neural control of a computer cursor or “keyboard” for communication.

Front 3

Major Types of BMI

Back 3

Invasive, Semi-invasive, Non-Invasive

Front 4

Subtypes

Back 4

Active, Reactive, Passive

Front 5

Parts of BMI/BCI

Back 5

Intention-Signal acqusition-preprocessing-feature extraction-pattern recogition - postprocessing - control signal - device - feedback

Front 6

Why should BMI work

Back 6

correlations between a precentral cell and specific arm muscles consistently appeared under several behavioral conditions

Front 7

BMI that stimulates

Back 7

cochlea

Front 8

HC design process

Back 8

Inspiration, ideation, implenation

Front 9

Three circles for design requirements

Back 9

Viable(business), Desireable(Human), Feasible(technology)

Front 10

dorsal

Back 10

Front 11

Regions of Cerebral Cortex

Back 11

Frontal, Temporal, Parietal, Occipital

Front 12

M1

Back 12

Primary motor

Front 13

SMA

Back 13

Supplemental motor

Front 14

Pre Motor

Back 14

PMd PMv

Front 15

Temporal

Back 15

Hearing, Memory

Front 16

Parietal

Back 16

Multisensory

Front 17

Occipial

Back 17

Vision

Front 18

Neural transmission input

Back 18

dendrites

Front 19

neuronal integration

Back 19

soma

Front 20

action potential generation

Back 20

axon hillock

Front 21

trasmission line

Back 21

axon

Front 22

site of information transfer

Back 22

synapse

Front 23

Postsynapic potentials

Back 23

excitaroy and inhibiatory

Front 24

Four steps spike sorting

Back 24

filter, spike detection, feature extraction, clustering

Front 25

filter

Back 25

high pass filtered raw voltage trace

Front 26

spike detection

Back 26

accomplished with a threshold that can automatically be set at say 3-5 standard deviations

Front 27

feature extraction

Back 27

there are a lot of possible feature spaces that one can use such as peak and valley of the waveforms from ii)PCA is often chosen 2D is shown but higer dimensions can be used

Front 28

clustering

Back 28

one method for extracting units template matching or the use of other distance metrics are others

Front 29

7 Techniques for recording from neuralactivity

Back 29

1.SingleSharp Electrodes (old school)●2.Arraysof Micro wires ●3.Arraysof Silicone probes 2D horizontal ●4.Arraysof Silicone probes 2D Vertical ●5.3DSilicone probes ●6.Non-electrodedesigns 1.Optical 2.Non-Optical

Front 30

ElectricalInterface

Back 30

•Purposeis to translate the change in ion concentrations due to neural activity intorecordable current/voltage changes.

Front 31

Half-cellpotential

Back 31

Thechemical reactions that govern the exchange of ions at the electrode–tissueinterface are given by C → C+ + e− and A− ← A + e−. Here, when the metal Ccontacts the electrolytic electrolyte fluid of the brain, the aforementionedreactions are triggered. In this example, the metal C is oxidized and gives offan electron whereas the anion A− gains an electron to be reduced. The chemicalreaction begins immediately and, as a result, local concentration of cations atthe surface of the electrode recording site changes.I?S'rV

Front 32

regresssion

Back 32

real valued output

Front 33

simple linear regression

Back 33

parametric statistical technique used to analyze experiments in which samples were drawn from populations characterized by a mean response that varies continoulsy with the magnitude of the treatment

Front 34

Linear Basis Function Model

Back 34

y(x,w)=w0+w1x1+...+wdxd

Front 35

linear regression model

Back 35

wiki