Class Notes (834,037)
Canada (508,290)
NESC 2130 (1)

Cog Psych Notes for Entire Semester

34 Pages
Unlock Document

NESC 2130
Olav Krigolson

Movement Planning and Motor Control: A) Article: Macar et al., 2004 : Timing functions of supplementary motor area: an event-related fMRI study -common set of brain areas is activated in relation to timing, irrespective of a difference in task features ->attention important for accurate timing-> right parietal and dorsolateral prefrontal cortex= sustained attention ->anterior cingulate cortex= supervisory attentional system -cerebellum, basal ganglia, and supplementary motor cortex (SA) have important role in timing 1) cerebellum: PET and fMRI studies show activated IN timing task, but activation not specific to timing processes; e.g. decisional/memory processes (indirectly important to timing) 2) basal ganglia: encodes temporal targets 3) SMA: explicit timing Goal: event-related fMRI study where subjects to produce a 2.5s target interval -predict stronger activation of basal ganglia and SMA in Time compared to force task (e.g. can compare to see what is activated for timing or for motor performance->both areas noted for motor use) Methods: -13 subjects -1 session= training; 2= fMRI recording -Time production and Force production task ->subjects held push-button, equipped with force transducer (measures force) in other hand ->respond by pushing button with tip of thumb -trials in each task began with presentation of different letter at centre of video screen (prompt) Training session -2 targets (interval and force) memorized in trial blocks which included accuracy constraints -order of time and force tasks counterbalanced Time Task -1st letter, ‘T’, appears on screen which remained visible for 1-1.6s -After ‘T’ replaced by ‘X’, subject had to press button 2.5s after ‘X’ appeared with optimal accuracy -then received auditory feedback indicating whether delay b/w presentation of ‘X’ and button press was correct, too short, or two long, much too short, or much too long Force Task -‘F’ appears which remained visible for 1-1.6s -‘F’ replaced by ‘X’, subjects had to press button with force of 8 N without any speed constraint -auditory feedback told them if correct, too weak, too strong, much too weak/strong -training in each task continued until 8 ‘correct’ responses in set of 10 tr4ials -then subjects performed another block of trails where Time and Force task were randomly mixed, until same performance criterion (8 correct) was reached Recording Session -subject in fMRI with force transducer -completed warm up trials and saw letters appear on screen ->e.g. after training session, no feedback after response, and letter ‘X’ remained visible until next trial ->letter used as fixation point to ensure identical visual perception during task and intervals -null trials used as baseline to compare with Results Time task-> activation in Caudal part of SMA, putamen and lateral cerebellum Force task-> left premotor cortex, right primary motor cortex, right middle temporal gyrus, left insula, and thalamus (assuming more in force b/c of all areas implicated in movement) Task specific activations: Time-> SMA proper and left primary motor cortex Force-> right primary sensorimotor cortex and left cerebellum, and infero-parietal cortex and insula Discussion Time task-> large activation of SMA (specifically proper) ->greater SMA activation in Time compared to Force can’t be explained by difference in length of execution period b/w two tasks -using other model-> stronger activation of SMA in time compared to Force -ERP, MEG, PET, and fMRI-> suggest SMA involved in programming operations (taking place before foreperiod) -confound: effect of response form on SMA involvement ->stronger response force produced in Time compared to Fork-> suggests motor features can be responsible for SMA activation -although putamen and cerebellum were activated, no specific contribution to process of either time or force; SMA was most prominent in activation - _____________________________________________________________________________________ B) Desmurget and Grafton, 2000: Forward modeling allows feedback control for fast reaching movement -3 debated models of how visually-directed movements are planned in advance or controlled during actual execution -feedback: process that mediates hand path corrections by comparing (1) target position and (2) an estimation of hand location 1) feedforward model: motor command is defined in advance of onset of movement ->role of feedback loops are minimal and at end of trajectory when hand velocity is slow 2) feedback model: pattern of muscle activation that is required to point to the target is defined during course of arm displacement, not prior to onset of movement ->no prior motor plan and muscle command is generated in real time, and can be updated if error ->think of thermostat-> compares current temperature to reference value then modulates responses to radiator 3) hybrid model: trade-off b/w 2 models -> imprecise motor plan assembled prior to onset of movement (feedforward) ->under constant ‘supervision’ of powerful internal feedback loops that adjust it in real time (feedback) -examine validity of the 3 models for reaching movements ->favour hybrid model-> preplanning and feedback are controlled by nervous system Sensory information and feedback control -over last 3 decades, feedforward models most influential ->through 1) model, sensory feedback thought to play marginal role in movement accuracy, b/c: a) somatic deafferentiation didn’t prevent subjects from executing accurate movements in dark b) some movements of short duration could be completed before minimum delay required to process sensory info c) online corrections that were based on sensory feedback could produce unstable trajectories for pointing that was performed at high or medium velocity *eventually ruled out that sensory feedback loops couldn’t be used to control hand trajectory -but goal-directed movements are more accurate when proprioceptive or visual info is present, so dual model proposed -dual model: reaching movements segmented into two components 1) driven entirely by motor plan and ensures rapid transport of hand to vicinity of target 2) depends on sensory feedback loops and allows corrections at end of trajectory, when movement velocity becomes low Feedback and need for forward modeling -forward model rationale: nervous system progressively learns to estimate behavior of motor plan in response to given command ->moderates/adjusts: by integrating info that is related to initial movement conditions, motour outflow and sensory inflow, probably position and velocity effector can be predicted -in other words: model receives input of copy of motor outflow -based on the info, end-point of movement can be predicted and continuously compared to target location -discrepancies cause error signal to be generated, which triggers a modulation of motor command Non-sequential, non-ballistic control of reaching -sequential control and ballistic reaching (think Woodsworth)= how motor plans are generated -non-sequential= coordination of eyes, head, and hand during goal-directed movement is chronological ->BUT EMG shows that eye, head and arm are nearly synchronous->therefore, motor command is sent to these effectors in parallel -challenge to ballistic movement-> imperfect estimate of target location correct early in course of movement, specifically during acceleration phase Feedforward specification of the motor command -b/c forward models adjust movements online, do any movement need to be planned in advance? -TMS study found: relatively accurate movements can be performed in absence of online feedback loops ->suggests motor plan is only crudely defined prior to onset of movements and subsequently updated through internal feedback loops as movement progresses ->loss of accuracy that is observed with feedback loops are disrupted support this (e.g. target moves during saccadic response) Functional Anatomy of internal error corrections -cerebellum associated with feedback control ->crucial in forward and inverse internal models Overall: no single algorithm can describe control processes used to perform goal-directed movements ->reaching towards a target requires a feed forward specification of motor command - forward modeling of arm - online updating of initial muscle pattern of activation are synthesized in reliable feedback loops -thought to involve cerebellum (and PPC) ______________________________________________________________________________ Lecture 1: Neurons as Detectors Lecture 1: Neurons and Neural Anatomy -cortex: small perimeter; brown/gray layer (1/4 inch thick) ->full of nerve cells->mostly neurons ->brain forms networks of neurons->connect via chemical/electrical signals -chess computer: to beat chess master 3.5 to 2.5 ->Deep blue: capable of evaluating 200 million positions but only narrowly won ->5 years to put together but human brain does chess and more! -brain: 2% of body mass -20% of 02 and blood supply -100 billion neurons; 0.15 quadrillion nerve synapses ->150,000 km of nerve cells Neurotransmission -stuff goes in dendrites -then to nucleus and axon -and either signal (action potential) goes or doesn’t ->neuron only fires if gets right info Synapse: gap b/w neurons ->axon generates electrical signal ->signal travels down axon (can do so b/c of myelin sheath) *Lou Gerigs disease= no myelin sheath 0 or 1-> neuron either fires or doesn’t ->at synapse: signal electrical ->at end, vesicle releases N.T. (chemical) ->when signal hits terminal, NT releases and goes across cleft and binds to receptors on dendrite side ->will create excitatory (EPSP) or inhibitory (IPSP) post-synaptic potential (chemical) -there are lots of dendrite connections: 1: 100000000000000000000000000000 -skin receptors: pressed/squished and then send EPSP/IPSP through dendrites to primary motor cortex ->somatosensory: gets signals from skin and at same time, brain says ‘hey I’m being stimulated!” -> so stimulate brain is same as being touched directly -behind eyes: rods and cones are receptors too -> photons hit these (R and C) and fire -> and then get activity in occipital cortex *know: -picture for each question Cerebellum Occipital lobe/cortex, parietal lobe, temporal lobe (dorsal/ventral stream) -primary motor cortex (M1) -Supplementary motor cortex- -lateral premotor area (LPA) -somatosensory -anterior cingulate cortex *know orientations, cuts *know about ataxia, and why man moves like that (has to rely on feedback) *know important names-> Wolport, Goodale, Woodsworth, Klapper and Erwin (76) *articles -gstds of neurons in visual cortex *1 picture -bar of light (yellow) is stimulus (in picture) -neurons will fire when requirements met/activated (e.g. stimulus in path of neurons) 1) is firing= yellow bar hits and neuron fires a lot 2) bar of light isn’t perfectly aligned to input->only some fire e.g. neurons could have be detectors for specific functions, like colour ->monkey’s neurons sensitive to colours in brain ->(could be colourblind if eyes affected OR brain/neurons are cortically damaged) *second picture -input layers of neurons doing multiple things ->light hits back of eye-> retina -retina: image grid, and one photon of light hitting that spot ->same spot/grid in visual cortex is stimulated -therefore, what hits eye is hitting back of visual cortex But how do you see the complex shape? (part 2: groups of neurons as detectors) Primary visual cortex (PVC): detects lines and bars PVC->next layer of neurons->etc. *next layer of neurons=add new layers of info -line detecting neuron-> L detector (detects that shape) fire b/c connected to line detectors ->Line detectors= stimulus hits retina which transmits to neurons which fire when see that stimulus To see square (complex shape): need bunch of line detectors ->all 4 L detectors fire, than square appears Many layers is how we see complex shapes: -dots->lines->L (angle)-> square-etc> ________________________________________________ Lecture 2: Motor Planning -monkey video: robotic arm and is motor on arm that makes it move -connected by wires to brain; neurons control movement ->controlled arm by though alone-no training ->how is monkey able to do that? Planning Movement -Primary motor cortex (blue strip): M1 -M1: controls muscles -motor cortex: elicits simple movements (e.g. twitch) in response to mild electrical stimulation -M1 strip=output-> i.e. part of strip specifically controls different muscles of body *stroke often affects motor strip, which is why affects movement so often -Lateral Premomtor Area (LPA): info starts in M1 , premotor, and LPA -> organizes direction of movement in space *image: each tiny line=neuron firing 0=monkey starts to move ->but see lots of firing before and after 0= neuron sensitive to direction ->and fires before starts to move (PREMOTOR AREA!) Supplementary Motor Area -involved in learning and executing action sequences -producing internally generated movements ->Marck Churchland article: Preparatory Activity in Premotor and Motor Cortex Reflects Speed of the Upcoming Reach *image: faster movements= bigger wave -called rostral neural firing graph Cerebellum -damage leads to motor deficits (e.g. ataxia: planning deficit) ->video: man w/ ataxia has hand sequencing issue (1) puts arm down and (2) stretch forward to get cup (2 stages not smooth; relies on feedback, no forward model) Alien Limb Syndrome -blockage of blood vessel leads to region of SMA (generates movements) and anterior cingulate cortex -unable to control body parts and feels they have own personality ->e.g. if a person passes reach out door, arm may reach out and grab knob Utilization Behaviour/ Environment Dependency Syndrome -disorder where objects in environment trigger individual to act, even if action is inappropriate ->objects trigger motor plan How are movements planned? Woodsworth (1989): every movement in 2 phases 1) Ballistic movement phase 2) Online control *for most, very fast/intertwined; w/ ataxia, clearly in 2 phases Movement Planning->How do you do ballistic movement? (2) Desired State (target location)-> want to put finger on tip of hand (3) inverse model (cerebellum)-> want to be at (2) and is at (1); provides trajectory to path (1) actual state (hand location)-> e.g. finger on arm ____________________________________ Lecture 3 a -inverse model: cerebellum generates path where wants to go (desired state) -LPA: direction -supplementary motor area: sequence of muscles So what is actually generated? Motor program: set of instructions *think motor memory ->abstract representation/ set of instructions, that when initiated results in production of coordinated movement ->issues commands to muscles (when, how, much, where to fire!) ->organizes muscles and joints into single unit Evidence for Motor Programs 1) Klapper & Erwin, 1976): -in one condition, people did simple movement (1 tap) -next condition: 2 taps *motor program takes longer to perform complex movements -next: 2 taps, touch nose -each gets more complex 1) looked at reaction time->from ‘go’ to start moving 2) response duration-> how long it takes -RT gets slower as task gets more complex Why does it take longer? -b/c takes longer for motor program to perform complex movement (*think more detailed instructions) 2) Woodman’s Results: -recorded EMG-> electrical activity on muscle -electrical activity begins before muscles moves -1 big peak: muscle starting to go up *motor program begins before arm moves -down peak: muscle stopping -extra waves: pointing more; corrective movements -slightly faded line: blocked movement ->initial muscle first fires (1 peak) -braking (stopped arm) movement also fired-> why brake blocked movement? -also saw corrective movements? -proof that brain sent signal to muscle LONG before muscle did it->proof of motor plan Invariant Parameters 1) order of events 2) phasing: temporal structure (relative timing->’how long does this fire for? 3) relative force *know difference b/w relative and absolute timing *picture: 3 muscles firing -squiggly line= firing ->relative timing-> if double speeds, divide all by 2 b/c relative: muscle ratios stay the same -relative force of letters stays the same (ratio) Variant Parameters 1) movement time-> can do it more quickly or less 2) movement amplitude-> how far/how hard movement is (e.g. how far to throw football) c) motor schema -schema: relationship b/w outcome (e.g. how far will throw ball) and amount of force -desired parameter: how much force needed -desired outcome: how far you want it to go -initial conditions: corrected for environment ->e.g. want to throw football 50m ->BUT is windy day ->initial condition= desired parameter corrected for wind -neuron: detects visual stimuli (e.g. line segment) -when in visual field, retina sends signal to visual cortex, where neurons with function of detecting that form of stimuli pick it up Group of neurons: to create whole image, multiple neurons may work in tandem on layers ->e.g. one set, picks up lines; other set connected to line neurons: curves -as a group: pick up full representation of image: circle (JP’s article) Lecture 4: Online Control -online control: what happens while you are moving Example: -try to serve volleyball and hit it in specific spot -why do you miss? -3 errors in motor command 1) memory isn’t perfect-> motor plan might be wrong 2) Neuromuscular noise even w/ perfect memory, still miss sometimes, why? -message gets sent to muscle has neuromotor noise (NMN) -NMN: excess electrical activity in system -what causes it?-> fatigue= excess neural activity 3)external factors->environment can change ->example: Goodale, Pelisson, and Prabalanc (1986): Target: O  : finger -when hear beep-> try to touch target -one trial: after beep, target movies Hypothesis: if target moved during movement AND planned on motor plan, would miss -results: people did not miss ->ocular saccade: eye jumps, not moves fluidly->you are blind during movement ->if you move target during saccade, you don’t know saccade (target?) moved -suggests people use online control-> can process sensory feedback ->pathway for perception goes through inferior temporal cortex ->another to parietal cortex (visual system for action; unconscious) -in essence: do get feedback from target AND able to act on it i.e. had actual state, but as move, actual state moves -> image difference b/w actual and desired state is error ->also, difference b/w 2 target locations= error -motor plan fires this b changing motor plan (online control) Woodsworth (1989): 1) Ballistic movement 2) Online control -(1) doesn’t need online control b/c just one smooth movement - visual processing= transmission; relay to processing regions of brain ->takes 100ms-300ms ->transmission time= lag, takes time, that’s why online control happens at end 1) Visual feedback: processes that soccer ball moved (from eye) 2) Goes to posterior parietal cortex 3) Say hey! Target moved 4) Cerebellum calculates new trajectory 5) Primary motor cortex makes new command Limitations of Closed Loop Control *can’t fix a) or b) b/c internal -issue w/ system: e.g. can still miss target if don’t move, why? ->speed-accuracy issue: faster= less accurate, but why? a) motor command might suck b) NMN Forward Models -must be way to fix this -evidence: people can fix trajectory at beginning of movement -but how if are technically blind from lag and saccade? ->fixed faster than feedback, why? Wolport: why can’t you tickle self? Answer: b/c of forward model->forward prediction (trajectory) desired end point error predicted movement endpoint -forward model has prediction, so cerebellum (inverse model) does backward thing -forward model does opposite (of inverse): generates motor command (efferance copy) ->copy sent to cerebellum ->cerebellum takes location and efference copy and predicts endpoint-> e.g. If I do this, where will I end up? (THIS IS AN ERROR) -so use forward model to fix motor command -don’t need visual feedback to do this->cerebellum allows to make first prediction -forward model does it entire time moving->constantly makes new predictions Wolport Study: a) one person: makes first -2 person: pushes fist down -result: fist moves, why? -force and grip overlap, no lag, so feedback automatic ->makes efferance copy and can predict load so can choose motor plan (i.e. forward mode:don’t move fist!) b) make first -now you try to push it -result: fist won’t move, why? -grip higher than load->lag in feedback , no move=no forward model *grip: force of hand holding it Load: pushing down 2 online control systems 1) process feedback 2) forward models *both make same solution -guy w/ ataxia has no forward model, so has to relay on feedback -forward model developed w/ practice -apraxia: inability to work w/ objects -identical apraxia: can’t pantomime using objects ->damage to parietal cortex****good exam q ___________________________________________________________________________________ Lecture 5: Vision Prosopagnosic: knows it, but can’t identify it ->recognize thing, but categories of things (e.g. animals, faces) are blurred -30 areas in brain for seeing-> e.g. movement, shape, form, etc ->2 parallel streams: 1) how->navigation 2) what-> recognize object (temporal lobe) -Philip (prosopagnosic) has damage to area near fusiform gyrus (detectors damaged) ->shows deficits in some pathways See world -back of eye= receptive field; neurons fire= see image -in eyes: bunch of neurons->dendrites sensitive to light ->lots of them: rod->black and white; night vision Cone-> more sensitive to coloured light *just a bunch of detectors -back of eye->midbrain->back of brain -midbrain: Pulvinar nucleus, lateral geniculate nucleus, superior colliculus= attention -primary visual cortex found in occipital lobe (back of head) ->PVC lights up when you see stuff Cortical Blindness -damage to striate cortex causes it ->blind in part of visual space contralateral to lesions -lots of type of detectors in visual cortex (e.g. lines, shapes, colours) ->e.g. neurons firing to specific angles or colours -PVC: think pixels= receptors waiting to get hit by light ->millions of receptors-> some fire to more complex things Example: -neuron fires when sees monkey face -fires more to basic shapes that resemble monkey fire -fires less to just some features found in monkey face Lots of visual areas-> each area= different type of visual representation ->e.g. MT area= motion (but not exactly) ->sees pictures moving very fast (like reel of film)=motion -V4= shape processing -V1= PVC -ventral stream: identifying objects ->vision: back of head-> through temporal cortex-> more forward, image gets more complex until=full image -dorsal stream: target location House Problem House= lines, angles, squares, colour, not MT (motion) ->all visual areas allow you to identify this -fMRI study: see what happens in brain when think a/b specific thoughts ->e,g, think hammer->lots of areas fire _____________________________________________________________________________________ 4. What is our “body image”, and what insight do phantom limbs provide about how our body image is formed? (ch 2) -sensory mapping begins in utereo; development ->points to plasticity of network maps -strip on cerebral cortex: penfield homunculus ->where different body parts are mapped onto surface of brain -phantom limbs=remapping of body part; remapping hypothesis ->e.g. phantom arm; arm is near face on cortical map, so when arm gone, sensory input from face invades territory previously vacated by hand (CH 3): Mirabelle born without arms; phantom limbs -when walks, phantom arms don’t move that much; straight -when talks, are very gesticular->gesticulation has different neural circuitry ->suggests that this circuitry stayed intact without any kinesthetic feedback IN HER LIFE ->implies we have hard-wired image of body and limbs at birth-an image that can survive indefinetly, even in face of contradictory information from senses Body Image - internally constructed ensemble of experiences- internal imagery and memory of one’s body in space and time ->to create and maintain it: parietal lobes combine info from many sources: muscles, joints, eyes, and other motor command centres -command: supplementary motor area (complex tasks) to motor cortex to get muscles to move ->identical copes of command signal are sent to two other major processing areas: cerebellum and parietal lobes- informing them of intended actions -so for phantom limb (body image): experience depends on signals from (1) remapping of sensory cortex, )2) info to move arm is sent to parietal info about body image ->convergence of info from 2 sources result in image of arm moving -also, consider visual feedback ->visual system seems arm not moving; command sent out again for arm to move; visual feedback returns informing that the arm isn’t moving= phantom paralysis ->fixed with mirrored box where amputated arm has illusion of having hand, and now with visual feedback, can control phantom limb ->body image/ control also depends on visual system (relaying signals) -mirror box not cure; once phantom arm taken out, and no visual, phantom pain could resume -genetic factors in determining body image: ->body image very malleable ->body image is a transitory internal construct that can be profoundly modified with just a few trick __________________________________________________________________ Exam questions Grading scheme: -content answering question -development of argument -references from Ramachandran or to 1+ original research articles -use of diagrams- -reference to relevant neural structure -grammar, brevity, original thought, etc 1. Neurons, and groups of neurons, can be thought of as detectors. Explain. Stimuli in visual field I: neuron -stimulus appears in visual field -photons hits photoreceptors (rod and cones; behind eyes) in retina -at this level, image a layer of neurons; photon hits specific spot on grid -the same spot on grid in visual cortex is stimulated *draw image of neuron grid grid -this spot, consists of neurons whose function is to fire when detects this type of stimulus; requirement met (e.g. photon light) Transition sentence: neurons don’t just work individually, they can function as a group II: group of neurons -groups of visual neurons can work together to represent complex stimuli -e.g. a circle -supported by Zwickel et al. (2007) network architecture model -individual neurons in visual cortex detect specific aspects of the circle; e.g. lines -layers of line detecting neurons firing-transmitting info -multiple layers of neurons converge through dorsal pathway to represent more complex feature: curved line->use info from line detection *image of grid with curved lines -layers of neurons combine into further in PVC to represent more complex representation of object ->line segments and curves lines create full, circular object -in total: multiple layers of neurons work in tandem -one set picks up line -other set connected to those line neurons pick up curves -as goup: pick up full representation of image: circle (Zwickel et al., 2007) ______________________________________________________________ 6. Differentiate the dorsal and ventral visual streams. Zwickel at al (2007) Ramachandran-ch 4 “new pathway= dorsal and ventra
More Less

Related notes for NESC 2130

Log In


Join OneClass

Access over 10 million pages of study
documents for 1.3 million courses.

Sign up

Join to view


By registering, I agree to the Terms and Privacy Policies
Already have an account?
Just a few more details

So we can recommend you notes for your school.

Reset Password

Please enter below the email address you registered with and we will send you a link to reset your password.

Add your courses

Get notes from the top students in your class.