Lecture 1 : Methods
Neuron doctrine ; The most relevant idea in neuroscience. We still consider neuron as the
fundamental unit in the brain. He came up with the idea of flow of information and suggested
synaptic integration.
- He was in conflict with Golgi, he thought the brain was more like a reticulum. Where the
neurons where physically connected.
Cajal took the Golgi stain (silver to isolate neurons), but he realized you need to use young brains to
get better resolution of how the neurons look. The neuron is the fundamental unit.
When you have the fundamental unit you can construct a system.
Maps -> Organize perception/sensorimotor activities (receptive
field/homunculus/hippocampus)
Networks -> Coordinated activity
Example system = visual system
Example Network = cortical column
The method you choose depends on:
- The question you have
- What you want to study
o Anatomy
o Function
o Computational processes
o Integration
1988 2015
,The idea is not one thing is more important or relevant than the other. It is the combination of the
techniques that gives you a more comprehensive view of the matter.
Research methods and experimental designs:
- Any experiment requires systematic manipulation of independent variables
- Dependent variable ; the outcome from the technique used
- Simple design ; The subtraction method -> identify the difference between two conditions
(to levels of an independent variable).
Example : Put someone in the EEG and present nothing (baseline) and then you present a stimulus.
You take the baseline and the stimulation and if you want to know what the stimulation does is the
difference between baseline and stimulation.
o Elaboration subtraction method ; Additive factor design -> involves several conditions
that increments progressively the differences between independent variables
Example : First, you create a baseline (fixation), then you give a passive word and then you have the
auditory system that is active. Then you make the patient say the word aloud and you can take that
difference. Then you make the subject create a verb and then you can subtract that from previous
conditions.
- You need so many subtractions because of the limitation of the technique. The spatial
precision is lower.
1. Factorial design -> systematic manipulation of two or more experimental
factors
a. For example look with ANOVA
2. Parametric design -> involves investigating how activation changes as some experimental
variable (parameter) is changed over a range of levels
3. Hillyard principle (EEG/fMRI) -> manipulate only the psychological/task condition of the
experiment, while keeping the physical stimuli constant
a. Example : Look at the phone or look at the stick. So the physical stimuli is the same
because you only point the attention to one or another, but both objects are always
present. You just ask to attend or not to attend. That allows you to see how the
cognitive effect works in the brain.
What is attention doing to visual perception?
Gives a cue and indicates attention to one or the other,
but they are both present
Studying the organization of the brain
- Behavioral methods
o Reaction time
o Accuracy
o Other (rating scale : likert-analogues)
- Psychophysics
o How subject will react after modulation of a cognitive factor
o ‘An exact theory of the functionally dependent relations of… the physical and the
psychology’ (Fechner 1860)
o Systematic relationship between perception and physical stimuli
o Weber’s law (Laws of perception) : ‘Magnitude of the stimulus increment must
increase in precise proportion to the stimulus already present, in order to bring an
equal increase of sensation’
, - Neuropsychology
o Former lesion-deficit studies
o Patient’s behavior – post mortem pathology (e.g. broca aphasia)
o The combined use of brain imaging has provided more accurate description of lesion
based studies (VLSM)
- Non-neural physiological responses
o Skin conductance, heart-rate, respiration rate, skin temperature, blood pressure etc.
o Eye-tracker
o Pupillometry
Examples of techniques
1. Electrophysiological recordings
(invasive/non-invasive ; EEG method) ->
patch clamp
- Membrane potential
o Problems ; not a good neuron to
record, pipe breaks etc.
2. Microwaves
- Tetrodes ; Allow single-cell activity identification. Unit isolation quality varies as a function of
distance from the electrode. multisite electrodes (a ware tetrode, for example) can estimate
the position of the recorded neurons by triangulation.
o The electrodes are sufficiently flexible and resistant and they can be placed in part of
the cortex or deeper and give stable recordings of the action potentials from the
extracellular side. E.g. how many action potentials per unit of time a neuron produces.
o Problem -> if you only use one electrode, how can you know that you are listening to
spikes of only one neuron instead of multiple? You can only see the multi-unit activity,
so the sum of all the activity.
, ▪ Improvement -> Use a multi-electrode recording in a configuration. You can
take the activity of one electrode and the other and plot it together. Because
every electrode has a bit of a different position the plotted shape of every
electrode will have a different shape. By combining the activities you can
separate the neurons
▪ The electrode we now use is a very long pin with different contact points and
you can also identify the location of the neuron and in which cortical layer.
3. Neuropixels (silicon & polimers-based electrodes)
- The contact points are now in an organization that allows you do to the same
recording as with tetrodes, but now with this only one point.
- Up to 700 neurons recorded simultaneously
- Problem -> Neurons in different areas have different maps and we don’t really know
how this differs in different levels. Now you can separate the different parts of the
brain and their recordings.
Neuropixels 2.0 -> spread the covering of the neurons you record
Local field potential :
- Captures all membrane fluctuations before you have a spike in the membrane of the spike. If
they are relatively synchronous it will create a field vector and that can be recorded by an
electrode that is not necessarily close to the source
o Membrane activity from a group of cells. Measurement depends on
▪ How many electrodes
▪ How close it is to the source
MEG ; Magnetic field of electric fluctuations. Better quality of the signal and allows you to record
certain areas better than EEG.
EEG ; Difference in voltage of electric fluctuations
4. Electrocorticogram array
- Essentially an EEG but over the surface of the brain
- Advantage ; all the layers that are between the electrode and the source are gone. You can
directly measure on the brain
Neuropixels allow you to record the LFP of local microcircuits. Here you have the LFP separated by
layers and identify massive fluctuations of electric currents in different layers. If you also have the
spikes you can identify everything really precisely.
5. EEG ; event-related potentials
- You compete the mean of many trials and you get traces that are related to the event that
was presented to the subject.
o Peak latency ; how much time does it take for the wave to appear
o Peak amplitude
Neuron doctrine ; The most relevant idea in neuroscience. We still consider neuron as the
fundamental unit in the brain. He came up with the idea of flow of information and suggested
synaptic integration.
- He was in conflict with Golgi, he thought the brain was more like a reticulum. Where the
neurons where physically connected.
Cajal took the Golgi stain (silver to isolate neurons), but he realized you need to use young brains to
get better resolution of how the neurons look. The neuron is the fundamental unit.
When you have the fundamental unit you can construct a system.
Maps -> Organize perception/sensorimotor activities (receptive
field/homunculus/hippocampus)
Networks -> Coordinated activity
Example system = visual system
Example Network = cortical column
The method you choose depends on:
- The question you have
- What you want to study
o Anatomy
o Function
o Computational processes
o Integration
1988 2015
,The idea is not one thing is more important or relevant than the other. It is the combination of the
techniques that gives you a more comprehensive view of the matter.
Research methods and experimental designs:
- Any experiment requires systematic manipulation of independent variables
- Dependent variable ; the outcome from the technique used
- Simple design ; The subtraction method -> identify the difference between two conditions
(to levels of an independent variable).
Example : Put someone in the EEG and present nothing (baseline) and then you present a stimulus.
You take the baseline and the stimulation and if you want to know what the stimulation does is the
difference between baseline and stimulation.
o Elaboration subtraction method ; Additive factor design -> involves several conditions
that increments progressively the differences between independent variables
Example : First, you create a baseline (fixation), then you give a passive word and then you have the
auditory system that is active. Then you make the patient say the word aloud and you can take that
difference. Then you make the subject create a verb and then you can subtract that from previous
conditions.
- You need so many subtractions because of the limitation of the technique. The spatial
precision is lower.
1. Factorial design -> systematic manipulation of two or more experimental
factors
a. For example look with ANOVA
2. Parametric design -> involves investigating how activation changes as some experimental
variable (parameter) is changed over a range of levels
3. Hillyard principle (EEG/fMRI) -> manipulate only the psychological/task condition of the
experiment, while keeping the physical stimuli constant
a. Example : Look at the phone or look at the stick. So the physical stimuli is the same
because you only point the attention to one or another, but both objects are always
present. You just ask to attend or not to attend. That allows you to see how the
cognitive effect works in the brain.
What is attention doing to visual perception?
Gives a cue and indicates attention to one or the other,
but they are both present
Studying the organization of the brain
- Behavioral methods
o Reaction time
o Accuracy
o Other (rating scale : likert-analogues)
- Psychophysics
o How subject will react after modulation of a cognitive factor
o ‘An exact theory of the functionally dependent relations of… the physical and the
psychology’ (Fechner 1860)
o Systematic relationship between perception and physical stimuli
o Weber’s law (Laws of perception) : ‘Magnitude of the stimulus increment must
increase in precise proportion to the stimulus already present, in order to bring an
equal increase of sensation’
, - Neuropsychology
o Former lesion-deficit studies
o Patient’s behavior – post mortem pathology (e.g. broca aphasia)
o The combined use of brain imaging has provided more accurate description of lesion
based studies (VLSM)
- Non-neural physiological responses
o Skin conductance, heart-rate, respiration rate, skin temperature, blood pressure etc.
o Eye-tracker
o Pupillometry
Examples of techniques
1. Electrophysiological recordings
(invasive/non-invasive ; EEG method) ->
patch clamp
- Membrane potential
o Problems ; not a good neuron to
record, pipe breaks etc.
2. Microwaves
- Tetrodes ; Allow single-cell activity identification. Unit isolation quality varies as a function of
distance from the electrode. multisite electrodes (a ware tetrode, for example) can estimate
the position of the recorded neurons by triangulation.
o The electrodes are sufficiently flexible and resistant and they can be placed in part of
the cortex or deeper and give stable recordings of the action potentials from the
extracellular side. E.g. how many action potentials per unit of time a neuron produces.
o Problem -> if you only use one electrode, how can you know that you are listening to
spikes of only one neuron instead of multiple? You can only see the multi-unit activity,
so the sum of all the activity.
, ▪ Improvement -> Use a multi-electrode recording in a configuration. You can
take the activity of one electrode and the other and plot it together. Because
every electrode has a bit of a different position the plotted shape of every
electrode will have a different shape. By combining the activities you can
separate the neurons
▪ The electrode we now use is a very long pin with different contact points and
you can also identify the location of the neuron and in which cortical layer.
3. Neuropixels (silicon & polimers-based electrodes)
- The contact points are now in an organization that allows you do to the same
recording as with tetrodes, but now with this only one point.
- Up to 700 neurons recorded simultaneously
- Problem -> Neurons in different areas have different maps and we don’t really know
how this differs in different levels. Now you can separate the different parts of the
brain and their recordings.
Neuropixels 2.0 -> spread the covering of the neurons you record
Local field potential :
- Captures all membrane fluctuations before you have a spike in the membrane of the spike. If
they are relatively synchronous it will create a field vector and that can be recorded by an
electrode that is not necessarily close to the source
o Membrane activity from a group of cells. Measurement depends on
▪ How many electrodes
▪ How close it is to the source
MEG ; Magnetic field of electric fluctuations. Better quality of the signal and allows you to record
certain areas better than EEG.
EEG ; Difference in voltage of electric fluctuations
4. Electrocorticogram array
- Essentially an EEG but over the surface of the brain
- Advantage ; all the layers that are between the electrode and the source are gone. You can
directly measure on the brain
Neuropixels allow you to record the LFP of local microcircuits. Here you have the LFP separated by
layers and identify massive fluctuations of electric currents in different layers. If you also have the
spikes you can identify everything really precisely.
5. EEG ; event-related potentials
- You compete the mean of many trials and you get traces that are related to the event that
was presented to the subject.
o Peak latency ; how much time does it take for the wave to appear
o Peak amplitude
