Neuroimaging Flashcards
X-ray- general description
- gives a course description of body
- x-ray tube fires x-ray, pass through body. opposite side of body is a film that captures the energy –> image
- x rays absorbed at different levels by different tissues
X-ray pro and cons
Cons: low quality/resolution
- single image doesn’t give much info on brain
- no info on moment-moment activity
Pros:idk lol
Cerebral Angiography
Contrast x-ray
- contrast = dye that absorbs X-rays at different amount (most common is iodine
- iodine absorbs X-rays at much higher level. it is injected into blood stream- gives you map of arteries in brain
- gives info about blood flow (stroke)
Pros and cons of angiogram
pros: done everywhere, with limited resources
cons: doesn’t show tissue, only vessels. not done for healthy people- risks
Computed Tomography (CT)
- variation of X-ray
- same idea: firing x-rays but they are received by digital receiver on the other side
- fired multiple x-rays instead of 1, and fire from multiple angles and compile it together
- lots of low resolution images, reconstruct into 1 better image
- 1970s
- relatively uncommon now
- improved images by -reconstruction
Pros and cons of CT
pro: improved x-ray image. first time separate difference btw fluid and brain. (valuable in stroke to see tissue loss)
- good for seeing large vesicle, lesions, brain size
cons: radioactive waves, not ideal for looking at brain in multiple sessions, exposure to radiation
- can’t see difference btw white and gray matter
- only as good as its algorithms (improved a lot since 70s)
MRI
- 1940s
- H atom has magnetic resonance (N and S pole), under strong magnetic conditions, H atoms align themselves
- then, perturb magnetic field, as they revert back, they release energy. measure this energy. different densities of H20 throughout body = different energies. energy measured by tube around you
ubc has 3 T and 7T (At 3T, it is 60,000x earth’s magnetic field)
MRI pros and cons
pro: amazing brain image, reconstruction of many slices = 3D image
- non invasive, no harm
cons:
Overlay plot
- pictures from multiple individuals who have similar symptoms
- find common areas that are damaged
- gives idea what regions are important for that loss of function
Diffusion Tensor Imaging (DTI)
- structural
- variant of MRI
- relies on how water molecule move in brain (algorithm)
- shows white matter
- water molecule in axons are limited to where they can move, algorithm looks for constrained water
pros and cons of DTI
pros: see white matter, in a number of conditions there is no obvious change in gray matter, but may see differences in white matter (ex: Arcuate fascicules: tracts of white matter that connect Broca’s are to Wernicke’s)
Electroencephalography (EEG)
- measure voltage changes on scalp, reflects neuronal activity
- usually 20-60 electrodes on scalp, any one can be picking up signals from thousands- millions of neurons
- course measurement of brain activity
Pros: fast, almost real time, easy, non-invasive, measure combo of muscle and brain activity
Cons: course measurement
describe neuron activity when going from awake from asleep
neurons fire less often, but more in synch with each other. greater amplitude and decreased frequency.
- delta waves = deep sleep, slow
- gamma waves = high frequency, working on intense mental activity
most intense brain waves
gamma in meditating monks woowwww
Positron Emission Tomography (PET)
- *functional
- uses radioactive ligand (i.e. cocaine, FDG) which will bind and decay -> positrons are emitted. they collide with electrons in the brain, and photons are produced, exit the head and are detected
- active areas of brain use more blood and thus have more radioactive labels
Cyclotron
creates radioactive molecule
FDG (fluorodeoxyglucose)
and cons to using it
- radioactive glucose
- brain areas that are more active need more glucose
- see stronger signal in these areas, gives baseline activity in brain
Application: in schiz the front of brain has less FDG than healthy
- indirect measure of brain activity
- scans are 45 min, how meaningful is 1 image to represent all that activity
- not good temporal resolution
- spatial isn’t great either
Mean difference images
Simulation-Control = Difference
- add up all individual difference images to get 1 mean
- individual does a task, and then control (rest)
issue: all individual differences are slightly different. sometimes average isn’t the same as any individual - done in PET and fMRI
PET cons
- expensive
- temporally slow
- poor spatial resolution
Pro:
useful for looking at specific systems (i.e. cocaine binds to areas involved in dopamine transmission, can see exactly where dopamine neurons are using pet- shown that dopamine system gets weaker throughout lifetime in drug addicts (all addictive
drugs))
- useful for looking at specific proteins (i.e. tau)
- useful for looking at lifespan/condition changes (stroke, maybe CTE)
Diaschisis
Loss of function in part of brain that normally gets input/output from a region that has been damaged
- seen with PET
fMRI
- early 1990s, discovered oxygenated and deoxygenated blood have different profiles
- as balance of deoxygenated vs oxygenated blood changes, you can measure that
- presumption: active regions need more oxygenated blood
- functional image laid on top of structural
== BOLD response (blood oxygen level dependent)
BOLD response: what is happening at the synapse?
- synapse is v active, needs energy therefore more oxygen
- oxygen comes from blood (astrocytes)
- astrocytes have glutamate receptors on them. when you activate these receptors, it causes dilation of blood vessels due to influx of Ca2+ allowing more oxygenated blood to the area
(more blood given than actually need)
Paired Image Subtraction
- 2 similar conditions (experimental and control)
- subtract control from experimental
- any overlapping brain activity is removed
- leave behind construct of interest
*quality of results will be directly related to quality of controls
Problems with interpreting fMRI studies (5)
- Spatial Averaging- (one trial you see tea pot, then you see it again. Presumed that brain activity would be same- maybe its not. Also spatial averaging over subjects- BUT all our brains are different sizes. So brains are squished into brain mold. Might run into epiphenomenon: you get an average result. But maybe that average doesn’t reflect any one individual. Analogy = books in your office on left wall and right wall, average book is in centre of room, but there are none in the middle
- Spatial resolution: under 3 mm for voxel= 3D equivalent of a pixel. Not capturing single brain cell activity
- Temporal resolution: At time 0, stimulus is presented (i.e. we see the teapot), we see BOLD response 6-7 seconds after. This means that we’re recording activity of things that happen earlier in time. Since brains are always active, this is a problem. At its tightest resolution, it is still on the order of seconds. Hard to distinguish bursts of activity within a second (we can get 500 or more AP a second)
- Not necessarily necessity: There are many brain areas that don’t necessarily show up on a BOLD that are very important. Many do, but are not as important as they seem. For example, language tasks show bilateral activity, yet we know that left hemisphere is mainly controlling language.
- Focus on increases in activity: the presumption is that if a brain area is important, It receives more oxygenated blood. BUT some areas just receive a lot of oxygenated blood all the time, i.e. the HPC has high level of oxygenated blood all time.
○ Some researches noticed, When patients went to relax, huge burst of BOLD. Probably because people were thinking a lot about other things, mind wanders. This lead to new approach….. Resting state functional connectivity MRI
The default mode network
- studied because of problem 5! ( some regions are more active at rest than during task)
- resting state functional connectivity MRI
- choose 1 region as seen region, every time activity in it goes up, see what other brain regions also do
- when BOLD goes down, see what else goes down
- no task given
default mode network:
○ Seen when mind wanders, not working on a task, thinking about what you want to do
○ Medial PFC, left and right inferior parietal, posterior cingulate (autobiographical thoughts), and medial temporal lobe (Sometimes)
Problems with interpreting fMRI studies (6-10)
- Regional hemodynamics: BOLD response graph isn’t universal. Looks different for certain brain areas.
- Confounds: anxiety, boredom: scanner is not a neutral space. Scanner is anxiety-provoking. Loud sound. If you are tired, you might sleep and twitch
- Confounds: drugs: i.e. coffee. Change how brain functions - shows up in BOLD in ways that are hard to predict
- Anticipatory hemodynamics: when you make someone do a task many times in a row, you start to anticipate the demands/what you will have to do. Brain anticipates what areas will need oxygenated blood, so they will get it before the regions are even active
- Reliability: main point: from one scan to another, you don’t get the same data. Recent meta-analysis did task and re-task, reliability was about 30%.
- Statistic: in a brain image, there are between 60,000 and 1,000,000 voxels depending on the scanner. Each one for your control vs experimental is basically a t-test. Normally, significance threshold is t= 0.05. (19/20 times your observation is correct). 5% of million is A LOT of false positives/negatives. It’s not stringent enough. To get around this is use a more stringent value. We’ve known this for years in fMRI, but didn’t do it (even today this happens in literature). A post-doc in Halifax put a salmon in a scanner, asked the salmon to look at pictures, rate emotions. When he performed the analysis. And during the emotional pictures there was emotional significance. (2011)
Stages of Consciousness
- Brain death: all functions of brain and brain stem have permanently ceased
- Coma: loss of consciousness is complete: cycles of wake and sleep disappear, eyes are closed. Usually temporary (2-4 weeks). Afterward, patients emerge into consciousness or one of the other states.
- Vegetative state: sleep-wake cycles, eyes may open to stimulation. Only behaviours displayed tent to be reflexive
- Minimally conscious state: patient may seem vegetative but shows signs of awareness, such as following a command, reaching for something.
Locked- in syndrome: (technically not a disorder of consciousness, because patients are fully conscious) but they cannot move and may be mistakenly deemed vegetative or minimally conscious. Many retain ability to blink/move eyes.
Summary of Adrian Owen’s paper
PET scan (Positron-emission tomography)
• Put girl in vegetative state in PET scan and showed pictures of family/friends to look for signs of response
• Saw her brain responded similarly to how healthy individuals
To answer what this meant, years of research
Covert Consciousness =
Hidden awareness in patients who are declared vegetative
Then, researches sedated group of healthy participants and exposed them to same combo of speech and non-speech sounds that elicited normal pattern of brain activity in vegetative patients.
- Healthy subjects were unconscious, and there speech perception part of brain were activated just as strongly as when conscious - This proved that "normal" responses to speech was unreliable indicator of covert consciousness
Next, they set out measuring brain activity while healthy participants did tasks like singing, walking, playing tennis. Mentally performing these tasks generates reliable pattern of activity similar to actually performing it.
- Used fMRI (which doesn’t require injection of tracer chemicals like PET does) discovered imagining playing tennis and walking room -> room in house elicited strong activity in premotor cortex (tennis) and parietal love and parahippocampal gyrus (room) - This technique worked when tried on first vegetative patient - Worked for 17% of patients tried - Suggest higher level of processing in vegetative patients than originally thought
also used EEG, but didn’t get good resolution. tasks were adaptive to produce activity on cortex
What now?
- Use of fMRI and EEG to detect awareness and communicate with non-responsive patients paves way for development of brain-computer interfaces
○ Translating specific thoughts into concepts for those with brain injuries
Also brings up ethical dilemnas: can you ask a patients If they want to continue living? Do they have the capactiy to answer that
