Neurobiosz - fMRI 2017.03.28.

Analysis of the fMRI data:
	> univariate activation: it was mostly used in the past
	> multivariate activation:
		> patterns of action
		> machine learning is used for image processing
	> Resting potential:
		> Turned out that many thing can be measured well, when the person whose brain is measured does not have a task
		> the brain is active during that time as well
		> brain activity is not random; however, its signal fluctuates
	> Task-based connectivity:
		> Finding correlation between the task and the brain activity
	>Fully (resting) connectivity:
Resting-state fMRI functional connectivity:
	> Functional connectivity (FC): finding interdependence of signals between distinct brain regions
	> Resting-state fMRI functional connectivity: if you ask the measured person not to do anything, some regions of the brain get activated
		> these regions controls many functions
		> it was thought to be an artifact
		> it is proven now that it is a real brain functional
Neural basis of Resting-state fMRI:
	> measured with invasive technics (electrodes)
	> direct demonstration that things measured with fMRI are real physiological processes
The human visual cortex:
	> Humans are particulary dependent on visual input
	> Sagittial Section:
		> corpus callosum in the middle
		> on the back there is the visual cortex (in the sulcus)
Functional Definition (for seeing):
	> Something moving up activates other cells than when that thing moves down
	> The is a kind of brain injury when the patient sees snapshots but cannot see motion
Visual cortex in macaques:
	> Monkeys are also particulary dependent on visual input as humans
	> Many studies are made on visual cortex of monkeys
	> We can identify low level areas > they recieves information from sensors (a single neuron is responsible for a small area in visual field)
	> Intermediate and high leve: they are responsible for large areas (e.g. half of the visual field)
		> Where? and What? is answered in different area of the brain
Patient DF:
	> They cannot copy images because they cannot understand image, but if you ask them to draw an apple, they can do it
	> Acting without percieving
	> Action vs Perception: visual output goes to two different places depending on whether information is percieved or triggers action
Retinotopical mapping:
Protocol for Retinotopy:
	> rotating a pattern and measuring what brain area is active
	> slide without title shows the result:  image of unfolded visual field > Human brain is retinotropically organized
Feature specific visual cortex areas
Human visual cortex:
	> We can localize regions that recognizes faces > if these areas are injured, patients cannot recognize faces
	> these areas can be localized easily with fMRI
	> Category-specific areas: it is not straight forward that they are responsible for recognizing a specific thing but they are more active when the task is to recognize that specific thing
Object recognition is very fast and efficient when viewing conditions are good:
	> Not only in monkeys but also in humans
	> within 200 ms, brain can process every important information
However...
	> Hard task the face recognition when images are closer to natural situations
Predictive coding theory:
	> brain makes predictions
	> predictions are made according to the environment
	> When the image is noisy, brain makes predictions iteratively > all predictions uses the previous predictions
Face processing in the human brain
	> Regions for face recognitions can be localized easily
Detorioration face images by adding phase noise
Combined use of fMRI and EEG:
	> it is difficult to say whether the picture shows a male or female
	> On the noisy images you need more technics to recognize
	> Performance decreases, processing time increases on both fMRI and EEG
	> P2 is specificly active in case of noisy images
	> If you combine noise and morf errors, fMRI shows intermediate stages (yellow)
	> combining EEG and fMRI helps to localize an image processing stage in both time (EEG) and space (fMRI)
resting-state fMRI function connectivity in the visual cortex:
	> repeated the same experience
	> Regions are localized that are particularly active
Active vision:
	> Helmholtz made interesting studies long ago
	> Visual cortex makes assumptions based on differences and previous experiences
	> Aloimonos discovered that even brain does not process all information, but only the relevant parts
	> On natural images most parts are irrelevant or redundant
How can we study that with fMRI??
Eye-movement pattern:
	> On the first image it is not told that the viewer should look at something, on the next image, it is told that the image shows a face
	> Distance between nose, mouth and eyes are important pieces of information
Attentional networks:
	> Executive functions: making decisions
Visual attention:
	> how the brain orients different visual inputs?
Top-down modulation:
	> Subpopulation with demaged brain (by stroke maybe?)
	> A large subpopulation ... (miről beszél ez az ember???)
Biased competition model:
	> There is a competition between stimuli
	> The winner inhibits the loser
Visual attention:
	Moving dots in two directions... and WHAT?
Shifting experiment from monkeys to humans in the recent years
Idegrendszeri megbetegedések:
	> Why is it important? > a lot of people is affected
Az emberi agy:
	> Huge difference between mammalian brains
	> Human brain is different mainly in frontal cortex > most problems are related with frontal lobe
Kognitív képességek változása felnőttkorban:
	> The human brain is best between 30 and 40
	> the population of Earth is aging, there is more and more old people every year > more and more interesting the brain diseases
MRI alkalmazási lehetőségei a transzlációs kutatásban:
	> Trying to use biomarkers: direct measurement
	> National Institution of Mental Health (NIMH): the goal is to understand the cause of disease and develop a cure according to that
Alzheimer kór: 
	> Researchers are looking for changes during aging
Default Mode hálózat:
	> We can design a drug that targets ...
Percieved pain:
	> Drug for pain
	> pain is a complex state > it has a direct component that goes up to brain, but the perception is affected by many things for example you are expecting pain or not, and also it depends on your mood
Neuroanatomy of pain processing:
	> Task is to try to develop to decrease the pain
		> considering a patient with chronic pain (he or she feels constant pain over years) 
		> the problem is that what we can do is that we ask the patient to score the pain on the scale 1 to 10 > it is subjective, not good for scientific measurements
	> the idea is the measure the activity that is always active when patient feels pain
		> Wise and Tracey: higher concentration of remifentanil causes to decrease pain region activity, but for example visual cortex is not effected
		> Continued to repeat these experiments: further studies failed even if the first experiments where really promising
Hogyan tovább?
	> Increasing the resolution
	> New method to map small delays to discovert intrinsic functional connectivities: Dynamic Time Warping instread of simple correlation
	> fMRI alapú neurofeedback: real-time fMRI control > allows you to use rewarding system