Lecture 5 - Neuroscience of Memory Flashcards
What are the cognitive neuroscience methods?
Human:
- neuroimaging (structural, functional)
- lesion (temporary, permanent)
“Cognitive neuroscience”
non-human animal:
- neuroimaging (structural, functional)
- lesion (temporary, permanent)
“behavioural neuroscience”
on the human side of things we will talk about neuroimaging and lesion work. We will asl otalk about these for animals but they are different from the techniques we use in humans.
we divide these techniques into cognitive neuroscience and behavioural neuroscience.
What is the structural neuroimaging we use in humans? What is the functional one? What is a key distinction?
Magnetic Resonance Imaging (MRI)
Functional Magnetic Resonance Imaging (fMRI)
a key distinction is that in MRI you are measuring the brain structure and in fMRI you are looking at funciton. YOu overlay the fMRI on the MRI to see structure and function.
What does MRI use and how strong is it?
Giant magnet ~ 10000 greater than earth’s magnetic field!
this is a magnet. We have to keep in mind that MRI scans are safe in principle but you have to be very careful when you go into them. You have to make sure you have no metal in your body, no metal in your accessories. This can lead to catastrophic consequences because anything metal will get sucked into the scanner.
What does cerebral spinal fluid look like in an MRI machine?
cerebral spinal fluid shows up as bright white.
this scan would be a coronal orientation. Theres lots of math that goes into generating these images. What we are doing in the scanner is taking pictures of slices fo your brain. We take pictures of the brainand then you put those slices together. Imagine taking a whole bunch of pictures and putting them together and this gives us a 3d image of the brain. We acquire the slices in a certain orientation but later we can look at them how we would like.
we want these images so we can look at similarities and differences between people, to understand what brains look like in general. We also do this to look for abnormalities like tumors etc.
What is DTI? What does it look like?
Diffusion Tensor Imaging
a whole bunch of bright colours in a web
this is a subtype of fmri
depending on how you set up your sequence in the scanner you get another image and do a lot of math on this image. Algorithms are used to color code different parts of this image. This technique looks at white matter tracks. The axions. It tells you which direction they are going.
Can be used to image the white matter
What can we use fMRI for?
Important for detecting and monitoring disease
often times we are mainly only certain that someone actually has alzheimers until they pass away and we can look at their brain but now there are some biomarkers that make us more certain
maybe someone has a head injury and they need a scan to see if there is any visible brain damage. We can also use this to monitor folks who have been diagnosed with a neurodegenerative disease like dementia. Now we are even doing it in healthy people and trying to see if anything predicts who will go on to develop alzheimers.
ALso, Important for understanding brain plasticity
Posterior hippocampal volume is larger in taxi drivers
her and her collaborators found that the size of the hippocampus was larger in taxi drivers than controls like bus drivers or non drivers etc. She found that the size was correlated with years of driving a taxi. This suggests that plasticity can change the structure of the brain. The most ideal version of this study would be scanning people before they became taxi drivers and following them along. These findings are typically specific to the posterior part of the hippocampus but an alternative hypothesis is that people with a larger hipocampus are drawn to careeers with spatial navigation and these people are going to last the longest driving a taxi. We can use brain scans to ask questions about learning induced plasticity. Other people have done research in this area and they have looked at the size of different parts of the cortex before and after musical training.
Important for demonstrating a “double dissociation,” namely when two areas are functionally dissociated by two cognitive functions, with each being affected to a lesion to one area but not to the other. Why is this useful?
another common use is to look for double dissociations. This means you might recall that we talked about brocas aphasia and wernickes aphasia. These are language deficits that arise from different areas of the brain. Wha happens in a double disossociation is you have lets say 2 different patients. The first patient has damage to this area of the brai nand the second has damage in another area of the brain. You can have this first patient having a problem with language comprehension but they have no problem producing it. They have damage to one part of the brai nand a loss of function A but not B. The other patient has the opposite problem, this is referred to as a double dissociation.
you can have a patient with damage to one area and they lose episodic memory. BUT they have no problem with semantic memory, no problem learning facts about their lives. You have another person with the opposite problem. This is a double dissociation. Researchers have used this to make arguments about brain behaviour relationships. Thisis often presented in a clean way but there is actually a lot of complicated things. EX: there are multiple partso f the brain that are connected to episodic memory so you can observe this in more than one area
what is alzheimers a form of?
dimentia
LOOK AT PICTURES ON SLIDE 9
what are some examples of different diseases that impact white matter?
MS, traumatic brain injury (trauma from sports related injuries, car accidents etc.) you can have shearing of the white matter tracts. You can’t see those differences on a stanndard MRI you reallt need the subtype of MRI to identify that. White matter is thought to be effective in alzheimers as well.
What is the most commonly used functional neuroimaging technology? What is another one?
The most commonly used functional neuroimaging technology is called functional magnetic resonance imaging (fMRI), which is based on measuring changes in blood flow. Another method, electroencephalography (EEG) is a method for measuring electrical activity in the brain. Here, electrodes are placed on the scalp.
there are different types. fMRI and EEG. We are not going to get into to much background but a general idea of how fMRI actually works. First, the brain is covered in blood vessels both on the outside and inner portions. This is really important for the use of fMRI.
What is the explanation of fMRI?
Oxygen gets delivered to neurons by haemoglobin
When neuronal activity increases, there is an increase in oxygenated haemoglobin but not all of it is absorbed by the neurons
Critically, this increase changes magnetic properties (which can be measured), hence the term blood oxygenation level dependent (BOLD) imaging
we are not measuring neural activity, we are measuring changes in blood flow. Would be asked this on a test. Oxygen gets delivered to neurons. When neural activity increases there is an increase in the oxygenated hemoglobin. You get oxygen coming to the neurons but whats really important to keep in mind is that more oxygen is delivered to the neurons than the neurons actually need. So what we are measuring is the change in oxyen. The increase in oxygenated hemoglobin changes the magnetic properties of the brain and we can use that for MRI
this is what it would look like. Person lies in the scanner and we give them task a vs task b. Try to see if brain is more activated when looking at cute puppies vs kittens and you would subtract the 2 to see the difference. You would calculate the percentage of change in activation. These images take a lot of math to develop. Those images do not show up on the screen instantly. Often images are at the group level as well because individual brain scans can be quite noisy. There are 2 subtypes of fmri task based and resting based. Resting based is you don’t get any tasks at all and you can observe resting state networks that are activated and communicating when you are resting with no specific task. People have criticized this notion of rest because people are probably mind wandering etc. THe other example is task based and that is the a vs b which is usually how it is done. You always want some baseline. We always want to know how much does brain activity increase from task a to be and the other way around. We might want an additional baseline as well because if we just subtract the 2 we might not know what regions have (i lost my thought). You have a baseline, taks a and taks b. You subtract each from each other and you also subtract each from the baseline. Interestingly because we are not measuring neuronal activity there is a sluggishness to it.
Look at slide 19!!
What do you know about the response time of fMRI?
the fMRI response, the changei n blood flow response is much slower so we only start to see changes as a function of the cute puppy 5 seconds later which poses limitations in th type of questions we can ask. If you are looking at something really rapis, fMRi will not be the best technique to use.
Neurons start firing within less than 1 sec (e.g., every 3 ms!)
Stimulus occurs at time zero.
fMRI response ~5 sec later!
What are the pros and cons of fMRI?
Pros:
Good spatial resolution
Safe
(its pretty good at allowing us to see the different subcompoenets of a region or a region from each other. )
Cons:
Indirect (not measuring neural activity)
Poor temporal resolution (this gets back to the sluggishness.)
Need to stay still!
(you have to stay very still because if you move the pictures get blurry so scanning babies would be very challenging. )
What are the fMRI analytic approaches?
- Examine local activation “blobology”
How active is an entire region in a given state?
(this is when we scan people, we do special math and we look at the blobs of the brain that are active. Taks a vs task b subtraction and we look at the blobs left over for task a )
- Examine “functional connectivity”
How connected are regions in a given state?
(we look at how active entire regions are for a given task. THe next technique is functional connectivity. Are regions actually talking to each other during the task? )
Examine “voxel-based patterns”
What is the pattern of activation in a given state?
(what we can do is zoom in on a given blob and we will notice that it is not a uniform blob, MRI images are made up of pixels but we don’t call them pixels we call them voxels because they are three dimensional. The colour coding is actualy a way to represent numbers. We can capitalize on this fact and see that there is a pattern. Imagine you are in an airplane and you look down at the city and some lights are really bright and some are very dim. This is what we want to know, We want to take all of those lights and levels of brightness and use them as a pattern to find if this patern in state A is the same as the pattern in state B. if you figure out the pattern for a specific state lets say for looking at puppies. Imagine you lose your ability to speak but you are put in MRI and are shown pictures of puppies. If the pattern shows up we can see if you can actually register the image still. this has implications for people in a coma etc. asking if people have consciousness when in a coma. )
what is functional connectivity?
Functional connectivity is determined by the amount of correlated neural activity in two brain areas
so you do a brain scan and you collect data for a given time period and you scan people for 10 mins. You have 10 mins of data and you extract the signal across that 10 mins and you can imagine that the activity in that region will ebb and flow. If it looks like the lower panel that is the 10 minute period. You you have uppy downiness and another region has upidowniness. They aren’t perfect but you can see they might have something in common. We would say that these are functionally connected .Mathematically you take the number associated with every time point, pull the number out, and you put that number in the excel spreadsheet for region a in a line and you put region b in the next coloumn. Then you just highlight them and perform a pearson correlation. You are calculating a correlation and if those two brain regions are correlated we say they are functionally connected ,they are communicating with each other. Region can be positvely or negatively correlated. E.g. negatively correlated when one region gets more activated the other becomes less activated.
What is the use of fMRI for learning and memory?
here is an example. Imagine she puts you in a scanner and asks you to think about your past. You do trials, you ask them 20 times in the scanner to think about the past. 20 trials of past and 20 trials of future. When we do that, when we compare the baseline task, we get this patter nof activation, and when we do the future task we get the other pattern of activation. Note this is a saggital orientation. What do you notice about these images? they are similar. They are not identical but they are really similar. This was a big discovery. When we imagine the past or imagine the future, the network that is activated is really similar. Researchers also noticed that when speaking to pateints with amnesia they had problems wit hremebering the past but also problems imagining the future.
● Example discovery: when remembering past events, individuals with “superior memory” showed increased activity in many different brain regions
○ prefrontal cortex, the hippocampus, and other regions
another exmaple comes from studying individuals with superior memory. (for example autobiographical memory). People who have really good memories activate certain parts of that network to a larger extent. Including the prefrontal cortex, the hippocampus, and some other regions.
are there brain regions that are active during remembering?
there are memory athletes people who train themselves. These patterns are a little different (some of those people might have a natural ability to remember). With practice many of us can get better and better. The folks she is referring to today are people who have this natural ability. AKA they did not train themselves. They identify as their whole life having a really good memory. So when you get them to lie in the scanner and think abotu their past, they activate that part more strongly than other parts do.
what is EEG?
the other technique is known as EEG. We spent some time talking about functional MRI and now we will talk about EEG. For this we are placing electrodes on the scalp. You are using this to infer what is happening at the level of neurons. Because the electrodes are placed on the scalp, this technique does not have very good spatial reoslutation but it gives you good temporal resolution. EEG is nice because it is not as critical that a participant stays still. You can do it in babies and still get a pretty good signal. Ideally you could use both techniques and that owuld give you a lot of rich info.
● Event-related potential (ERP): electroencephalograms (EEGs) from a single individual are averaged over many repetitions of an event (i.e., repeated stimulus presentation)
● Compared with fMRI, EEG is a simple and inexpensive way to monitor changes in brain activity
(in EEG, we average across trials, when taking eeg data from a single individual you will average many repetitions of the same event.)
What is recording from neurons? Is it used in human or comparative neuro?
Comparative
Recording from Neurons
● Memory is affected not only by which neurons fire but also by how often the neurons fire
● Neurophysiology: study of the activity and function of neurons
● Example: measure the firing patterns of individual neurons is single-
cell recording (using an implanted electrode to detect electrical activity (spiking) in a single cell (such as a neuron))
(memory is affected not only by which memories fire but also how often they fire. Ex: you can measure firing patterns of individual neurons called single cell recording. There is a similar technique in humans but the only way it is done is when patients are preparing for surgery. Before the surgery is performed they monitor activity from neurons in parts of the brain so they are careful about which ones they remove. )
(Monkey moving stick task) : the task here is quite simple. The monkey moves the stikc arounf and depending on the orientation, the neuron being monitored wil fire differently. There is lots of firing that happens for orientation 1 (when the animal ispushing the stick forward. We know different neurons are different in theri firing for different orientations.
LOOK AT SLIDE 30
Can researchrs change brain activity? Does this happen in human or comparative neuro?
comparative
● Electromagnetic stimulation of neurons
○ Researchers can use microelectrodes to stimulate neural activity by delivering tiny
amounts of electrical current into the brain
○ Used as early as the 1800s to show that neural activity in the motor cortex produces motor behavior
this is how they mapped out the hemunculous on the motor cortex.
How does electromagnetic stimulation of neurons work?
● Electrical stimulation can lead to changes in cognition
○ Transcranial magnetic stimulation (TMS) changes activity in the cerebral cortex by generating strong magnetic pulses over the skull
there are more modern techniques that get at the same idea. TMS this uses strong magnetic pulses delivered over the skull and changes patterns of functional connectivity aka it changes the way the neurons are talking to each other. Because TMS is changing functional connectivity and it can be localized, the functional changes you observe are different when placing the coil in different places.
this technique is also used therapeutically for severe forms of depression
people might use this to map functions on different networks.
How do researchers change brain activity using biochemical control of brain states?
● Another method for manipulating neural activity is the use of drugs. These are chemical substances that alter the biochemical functioning
○ Drugs that affect the brain typically change neural activity by altering synaptic transmission
another technique you can use is drugs, pharmecological technique. Drugs can be adminsitered that cross the blood brain barrier and effect neuronal transmission, THey can change the release of a neurotransmitter or can control the reuptake ofthe neurotransmitter. There are some pharmacological techniques that work well for a lot of people, and although many people benfit from them, we don’t fully understand the mechanism by which they work. We have treatments that work but we don’t understand how. So for folks with depression there are different treatment options available.
are there drugs for learning and memory?
● Only few drugs have been developed to affect learning and memory
● More common are drugs that affect memory as a side effect (usually
negative)
