Flashcards in week 5- TMS Deck (53):
Background of lesion studies and classic cases
-A lesion is said to have occurred when a localised region of the brain is damaged and ceases to function normally.
-By measuring what the subject can no longer do, you -can then deduce what function the lesioned area of the brain was responsible for.
-Classic examples of lesion studies include
Henry Molaison (HM)
problems for studying lesions after they occur
-The researcher had to wait for lesions to occur (small n)
-The researcher did not get to decide what was lesioned.
-Often, had to wait for the subject to be die before the true extent of the lesion could be precisely determined.
-Given time the brain would often compensate and the subject would (at least partially) recover – from that point on the researcher is then studying an abnormal brain.
-Many of these objections can be avoided by studying animals, but if you want to study humans, you need a better technique.
It allows you to perform reversible lesions.
Thus, people are actually willing to participate.
The researcher can decide where the lesion occurs.
The subject can be tested before the brain compensates.
what does TMS do?
Uses a rapidly changing magnetic field to induce electrical currents in the brain
Sometimes this excites the brain (e.g. stimulation of the motor cortex can generate movement ticks)
More usually disrupts or hinders processing in that brain area creating a “virtual lesion”
The lesion is short-lived and can be made to occur at a precise place at a precise time.
-The first successful TMS experiment was by “The -Sheffield group” – Reza Jalinous, Ian Freeston and Tony -Stimulated the motor cortex using TMS
-Observed muscle twitches
Barker - 1985
TMS is based on Faraday’s principle of electromagnetic induction.
In brief, a changing magnetic field will cause a current to flow in a wire that passes through the magnetic field
This is how dynamos, electrical generators and electrical transformers work
basic principles continued
A current flows around the TMS coil generating a brief (approx 1ms), very strong (2 tesla) magnetic field
2 tesla ≈ 40,000 times the earth’s magnetic field
Because the scalp is permeable to the magnetic field, the magnetic field penetrates the brain, inducing a brief electric current.
The electric current is usually confined to the upper layers of the brain (i.e. the cortex)
How To Deliver Such a Large Pulse
To generate such a large magnetic field requires a lot of energy
A TMS machine works by charging up a capacitor and then suddenly discharging it to create the current pulse required to generate the magnetic field
Types of Coil
For brain stimulation, one often uses a butterfly or figure-of-eight coil
Generates a more focused current
Each coil generates a magnetic field in the opposite direction, thereby generating offset current loops that also circulate in opposite directions
Thus, a large localized current is generated
Generative Effects of TMS
-When applied over the primary motor cortex, can produce an observable twitch
-When applied to visual cortex TMS can generate phosphenes
-Phosphenes are brief flashes of light perceived in the absence of any visual stimulus
-The pulse has to be very strong to elicit a phosphene
Inhibitory Effects of TMS
creates virtual lesions:Thus, if the brain region was critical to the particular task that the subject was performing, then performance on the task would be reduced.
-Essentially, the TMS-induced current causes neurons to fire randomly, increasing the level of neural noise, thereby masking the neurons that are firing correctly
-These means that an areas cease to function correctly, though usually the processing is usually not totally disrupted
Advantages Of TMS Over Real Lesions
-It produces a focused “lesion” where and when you want it
-Does not give the brain time to reorganize
-Does not give the subject time to learn compensatory behaviours
-Because the TMS “lesions” are short-lived, each subjects serves as his own control – the perfect match. One simply has the subject perform the same task twice – once with TMS and once without – and compare his/her performance in the two conditions.
Why Not Just Do Neuroimaging (e.g. fMRI)?
-just because a brain area increases its activity when a subject performs a given task does not mean that it is essential to the performance of the task – it might be epiphenomenal
-TMS can be used to deactivate those brain areas “found” by an fMRI scan to investigate whether they really do play an essential role in the task being studied.
-TMS can also be used to determine WHEN a given brain area plays a role in a given task
-the brain structures essential to life (e.g. the pons which helps to control respiration rate) are located deep within the brain out of the range of TMS
-However, TMS can still cause seizures
-Single pulse TMS is considered very safe
-However, rTMS (repetitive TMS) is more likely to cause seizures
-Strict guidelines (Wasserman guidelines) are used to prevent seizures in rTMS.
Clinical Use of TMS
The Royal Australian & New Zealand College of Psychiatrists (position statement 79, October 2013) has endorsed TMS as a treatment option for depression.
This option is available in Victoria (e.g. the Victoria Clinic)
Typically this is a treatment of a last resort when all other treatments have failed
Perhaps still a bit controversial but getting increasingly accepted
Amassian et al. (1989)
-Amassian et al (1989) was one of the first studies to use -TMS to inhibit processing in a brain area
-“Virtual lesion” mode – the way TMS is typically used today
-Specifically, they used used TMS to mask a visual stimulus
-The stimulus was a trigram
-Each trigram contained 3 letters
-After a delay of 0-200ms from the presentation of the letters, a single TMS pulse was administered to the visual cortex, i.e. the rear of the head
Data for one subject
Control was when no TMS pulse was applied
Incidentally, no subjects reported any phosphenes (pulse was not strong enough to induce phosphenes)
Thus, masking not due to phosphenes
Possible objection: The TMS can produce various non-specific effects that might account for the results.
how is the visual cortex organised?
The visual cortex is retinotopically organized.
Each part of it processes input from just one region of visual space
However, the mapping is inverted up/down and left/right
Amassian et al. (1989) control experiment
-Made use of this retinotopic organization in their control experiment
-They shifted the TMS coil from left to right thereby moving the virtual lesion
-They reasoned that when the “lesion” is removed from the the part of the visual cortex that process a particular letter, the letter should become visible.
-Their results were consistent with this
-As before, trigrams horizontally oriented. Delay between presentation of trigram and pulse was 100ms
Amassian moving tms coil left
Left letter processed in right visual cortex
Thus, the further left the TMS coil is moved, the less suppression of right visual cortex, the less suppression of the letter, the higher the accuracy for reporting that letter
A beautiful within-subject control for non-specific effects of TMS
amssian moved to the right
Note, that when TMS coil moved far enough to the right, reduction in suppression also occurred, further evidence that one needs to suppress the area responsible for processing that letter, not just the relevant cerebral hemisphere
Amassian follow up experiment
-In a follow-up experiment, arranged letters vertically and moved the coil rostral (up)
-Effectively, obtained the same result
the delay between the presentation of the letters and the pulse was 100 ms
-Top letter processed by lower visual cortex
-As move coil up (i.e. rostral), less suppression of lower visual cortex, so less suppression of letter, so higher accuracy
Amassian take home message
-used TMS to create a “virtual lesion” in the primary visual cortex (V1).
-The lesion prevented the processing of the letter that -otherwise would have been processed in that area
-The lesion was highly localized in both time and space
Amassian et al. (1993)
-Considered a situation where two trigrams were presented in quick succession.
-Normally, the second trigram would prevent the first from being seen.
-The first trigram is said to have been masked
Amassian et al 1993 2
-investigated whether they could use TMS to suppress the representation of the second trigram thereby allowing the first trigram to be seen
-this study they aimed to use TMS to create a positive effect (causing a letter to be seen)
-TMS inhibits the perception of the second trigram allowing the first trigram to be perceived.
Amassian hypothsis 1: noise hypothesis
-that the second trigram suppresses the first trigram by introducing noise into the primary visual cortex
-This noise prevents the first trigram from being seen
-If true, then introducing more noise into the primary visual cortex should mask the first trigram even more
-Because TMS increases the noise in the visual cortex…
…we would expect a TMS pulse to further mask the first trigram.
-In fact, it unmasked the first trigram…causing us to reject this hypothesis.
Amassian hypothesis 2
-A more plausible hypothesis is that the interactions between the two trigrams occur beyond the primary visual cortex
-In other words, this experiment shows that masking occurs not in the primary visual cortex (as had previously been thought) but elsewhere in the brain
-Application of the TMS thus prevents the second trigram from progressing beyond the primary visual cortex and thus masking the first trigram
-Without TMS, representations of both trigrams progress beyond the primary visual cortex
-Consequently the 2nd trigram suppresses the first one
-With TMS, the representation of the second trigram does not progress beyond the primary visual cortex
-Consequently, the 1st trigram is no longer suppressed
take home message amassian et al 1993
-TMS can generate a positive result (i.e. it allowed the first trigram to be seen)
-This is useful because Amassian et al. (1989) had shown the opposite – i.e. how TMS can be used to suppress the perception of a trigram, a negative result.
-Taken together, these two studies demonstrate that TMS can have both positive and negative effects
does masking occur in primary visual cortex?
-masking does not occur in the primary visual cortex but -must occur at a later stage in the visual processing stream.
-We could not have determined this without using TMS because we needed to create a lesion at a precise time
-Conventional lesioning methods (e.g. surgery) cannot do this
Where in the brain does visual awareness emerge?
Hypothesis: The further along the visual processing pathway the more the neural activity will correlate with visual awareness
Middle Temporal Area (MT)
-MT seems to be associated with motion perception.
-Activity in MT correlates well with perceived motion.
-Lesioning MT causes akinetopsia – the inability to perceive motion.
-Simulating MT causes spurious motion perception
-However, neural activity in MT does not correlate well with other aspects of visual awareness such as colour.
is there an area for all of visual awareness?
There is no one area whose activity correlates with all aspects of visual awareness.
The Master Map Hypothesis
to form a coherent object, We need to relate together information from different brain areas.
The brain area responsible for the master map must have a high resolution (because visual awareness has a high resolution).
V1 fits the bill
master map continued
The detailed processing is carried out elsewhere (e.g. motion processing occurs in MT, colour processing occurs in V4)……but then all this information propagates back to V1……allowing V1 to relate it all together……thereby creating a coherent perception…
allowing you to see a single coherent object.
example of master map
For example, suppose the subject is shown a moving red disk.
V1 would register the location of the disk (it can do this because it has a high spatial resolution).
MT analyses the disk’s motion and reports this information back to V1.
V4 analyses the disk’s colour and reports back this information to V1.
V1 relates these two bits of information together.
This is what allows us to be aware of both the disk’s redness and movement
Notice that neither MT nor V4 could deduce that the disk was both moving and red because each analyses only one aspect of the disk (MT analyses its motion, V4 its colour).
This is why V1 was needed.
what did Pascual-Leone & Walsh (2001) test?
Prediction: Deactivating V1 should prevent visual awareness from occurring even when later visual areas (e.g. MT) are active.
what did Pascual-Leone & Walsh (2001) do?
-They used TMS to stimulate MT – subjects reported seeing moving phosphenes.
-Then, they again used TMS to stimulate MT while simultaneously using TMS to inhibit V1. Now, no moving phosphenes were seen.
-In the main experiment, the TMS pulse applied to MT/V5 was at 100% threshold for producing a phosphene
-The pulse applied to V1 was only at 80% threshold for producing a phosphene – thus no phosphene was produced in V1.
-However, the V1 TMS pulse would be able to mask the phosphene from MT (Amassian et al. (1989))
-When V1 TMS pulse applied 5-45ms after MT TMS pulse, visual awareness of phosphene reduced (i.e. interference observed)
-No such interference observed when V1 pulse occurred ahead of MT pulse
results Pascual-Leone & Walsh (2001) ?
-The very short latency required for maximum interference is consistent with findings that the projection from MT to V1 has a very short latency
-Results suggest that unless MT activity is represented in V1, movement is not perceived
-Demonstrate the importance of the backprojections of MT to V1 in the awareness of motion
-Master map hypothesis supported
possible objection to findings of Pascual- Leone & walsh
-Perhaps the TMS pulse in V1 created a phosphene that then propagated up to MT and masked the activity in MT
-This seems unlikely because
The TMS pulse in V1 was subthreshold so could not have produced a phosphene
If V1 activity was required to propagate from V1 to MT, then one would expect suppression to be greatest when the V1 pulse preceded the MT pulse
Pascuel Leone & Walsh control experiment
-A second possible concern was that the inhibition might be due to non-specific effects of TMS (e.g. the loud click produced by the pulse)
-To control for this possibility the authors performed a control experiment in which they delivered both pulses (i.e. a subthreshold pulse and a suprathreshold pulse to MT)
-Because they were not inhibiting V1, they predicted that the phosphene caused by the stimulation in MT should now be observed.
-Conversely, if the previously observed inhibition had been due to non-specific effects of TMS then inhibition should still have been observed.
-In practice no inhibition was observed, ruling out non-specific effects of TMS
Pasuel Leone and walsh take home message
P-L & W used TMS to stimulate MT directly without the stimulation first having to travel through V1
This allowed them to ask whether stimulation of V1 was necessary for visual awareness
When they prevented the stimulation entering V1 from MT, awareness was abolished indicating that V1 is necessary for visual awareness.
Ashbridge et al. (1997)
Two conditions: a feature search condition and a conjunction search condition.
Regardless of the condition, each display always contained 8 items.
In the feature search condition, the target was a green vertical bar and the distractors were green horizontal bars
The target contained a feature (verticalness) that the distractors did not have.
Thus, the participants could search for the target by looking for the occurrence of this feature.
Asbridge conjunction condition
-In the conjuction condition the target was again a vertical green bar.
-But now some of the distractors were horizontal green bars and some where vertical blue bars.
-The participants couldn’t find the target just by looking for a single feature.
-The target was not the only vertical item
-The target was not the only green item
Thus, to find the target, the participants would need to search for a particular combination of features – green and vertical.
asbridge bottom line
-Feature search is predicted not to require attention (i.e. be pre-attentive)
-Conjunction search is predicted to require attention
Thus, if we could use TMS to disable attention, -performance on feature search tasks should not be affected but performance on conjunction search tasks should be adversely affected
-Ashbridge et al. (1997) tested this hypothesis and found it to be true
what part of the brain is involved in attention?
The right superior parietal cortex is thought to be involved in the deployment of attention
attention inhibited in Ashbridge
-Subjects were presented with the stimulus and required to indicate whether the target was present or not as quickly as possible
-By applying TMS to the right superior parietal cortex attention could be inhibited
-First, TMS was not applied and the average time taken to make a response was measured
-Then TMS was applied to the right superior parietal cortex and the experiment was repeated
attention inbited continued
-For each subject, the reaction time in the TMS condition was normalized by the reaction time in the non-TMS condition
-if the normalized RT was greater than 1, this means applying a TMS pulse increased the reaction time.
- inhibiting attention made visual search harder
results for feature search ashbrdige et al.
-Hindering the deployment of attention had no effect on feature search
-This is consistent with the hypothesis that feature search occurs in the absence of attention
results for conjuction search
-Conversely, hindering the deployment of attention had a large effect on conjunction search consistent with the hypothesis that conjunction search requires attention to be deployed
-Disruption was greatest at a delay of about 100 ms
-attention is required for conjunction search