Creating Better Learners by Driving Neuroplasticity

  • 976 views
Uploaded on

The CERI OECD/National Science Foundation International Conference took place in Paris, at the OECD Headquarters on 23-24 January 2012. Here the presentation of Session 1, Brain, Plasticity, and …

The CERI OECD/National Science Foundation International Conference took place in Paris, at the OECD Headquarters on 23-24 January 2012. Here the presentation of Session 1, Brain, Plasticity, and Learning, Item 2.

More in: Education , Technology
  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
    Be the first to comment
No Downloads

Views

Total Views
976
On Slideshare
0
From Embeds
0
Number of Embeds
0

Actions

Shares
Downloads
24
Comments
0
Likes
1

Embeds 0

No embeds

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
    No notes for slide

Transcript

  • 1. Améliorer les performances des apprenants grâce à la plasticité cérébrale Prof. Andrea A. Chiba University of California, San DiegoTemporal Dynamics of Learning OECD 2012
  • 2. Temporal Dynamics of Learning OECD 2012
  • 3. The goal of our Center:To understand the role of time and timing inlearning From the neuronal and millisecond scale… …to the brain and year-long scaleTemporal Dynamics of Learning OECD 2012
  • 4. Time is an organizing principle of the world Le temps est un principe organisateur du monde Ο χρόνος είναι µια οργανωτική αρχή του κόσµου•  Sequences underlie much of what we learn•  The brain is organized 5+3x6=48 =oops! according to temporal patterns as is our world•  The ability of the brain and the world to integrate is crucial THE CHILD JUMPED ON THE FROG. THE FROG JUMPED ON THE CHILD. Temporal Dynamics of Learning OECD 2012
  • 5. Time is an organizing principle of the brain STDP Spikes Local Field Potentials Cortical EEGTemporal Dynamics of Learning OECD 2012
  • 6. How are temporal sequences organized? Temporal pattern & sequence wi Elements A B C D Dt Dt Dt 1 2 3 time  ● Temporal patterns are learned and recognized ● Temporal pattern recognition is basic for perception, language, and cognitive processing ● Deficiency in rapid temporal processing underlies some learning deficitsTemporal Dynamics of Learning OECDInitiative 1 2012
  • 7. How do brain circuits learn sequences? Naive modelTemporal Dynamics of Learning OECD 2012 OECD 2012
  • 8. Time matters for processing… say Frequency stay 100 ms Time (milliseconds)These wave forms are identical except for the artificiallyinserted gap and a compensating shrinkage of the waveform.Temporal Dynamics of Learning OECD 2012
  • 9. Time matters for processing… say Frequency stay 100 ms Time (milliseconds)These wave forms are identical except for the artificiallyinserted gap and a compensating shrinkage of the waveform.Temporal Dynamics of Learning OECD 2012
  • 10. Time matters for learning… 100 95 90 85Percent Correct SYLLABLE Tone Duration = 80 RANGE 75 msec 75 Tone 1 = 100 Hz, PHONEME Tone 2 = 300 Hz 70 Control RANGE 65 Language Impaired < 40ms - Phonemes 60 40-350ms - Syllables 55 50 150 305 428 947 1466 1985 3023 3543 4062 8 15 30 60 Tallal & Piercy (1973) Nature. Interstimulus Interval (ISI) in milliseconds 6-8 year old language impaired children can t do the task at short ISI s: correlation, or causation? Temporal Dynamics of Learning OECD 2012
  • 11. Timescales of investigation Ba   Da   Whisking10s of msec Spikes STDP Phoneme processing Neuromodulation Motor action 100s of msec Perceptual decisions Brain rhythms Action selection Seconds Prediction Temporal Diff. learning Reinforcement Minutes LTP Days Neurogenesis Weeks Spacing effects Years ExpertiseTemporal Dynamics of Learning 11 OECD 2012
  • 12. Clarke  County  School  District  •  The  Clarke  County  School  District  of  GA  implemented  the  Fast  ForWord   products  in  eight  schools  during  the  2006-­‐2007  and  2007-­‐2008  school  years.  •  2,257  students  had  their  reading  achievement  assessed  with  the  Criterion-­‐ Referenced  Competency  Tests  (CRCT).  Some  students  used  Fast  ForWord   products,  others  served  as  a  comparison  group.  •  Based  on  CRCT  results,  on  average,  the  Fast  ForWord  parMcipants  made   greater  gains  than  the  comparison  group.    •  Following  Fast  ForWord  parMcipaMon  in  2006-­‐2007,  40%  of  students  who   were  not  proficient  in  2006  crossed  the  proficiency  threshold  in  2007,   compared  to  27%  of  the  students  in  the  comparison  group.  These  results   were  replicated  by  students  who  first  used  Fast  ForWord  products  during   the  2007-­‐2008  school  year:  42%  of  the  parMcipants  crossed  the  proficiency   threshold  in  2008,  compared  to  29%  of  the  students  in  the  comparison   group.  Temporal Dynamics of Learning OECD 2012
  • 13. Clarke County School District: CRCT Results 806 804 Proficiency Threshold 802 Scale Score 800 798 796 794 792 2 0 2006 2007 2008 CRCT Wave 1- Started Fast ForWord in 2006-2007 Wave 2- Started Fast ForWord in 2007-2008 Wave 3 - Did not start Fast ForWord before 2008 CRCT m08clrk03mf longitudinal scaled score CRCT wave 1 n=571 wave 2 n=758 wave 3 n=928Temporal Dynamics of Learning OECD 2012
  • 14. Brain Activation Improves Stimulus Fidelity in the Cortex ● Rodent Audition (Harris): Neural encoding of auditory temporal patterns is highly nonlinear and depends on brain state. Inactive brain state synchronized EEG, sleep, anesthesia Active brain state desynchronized EEG, alertness, attention State for effective coding and learning of temporal patterns? From the Harris Lab AM noise stimulusTemporal Dynamics of Learning OECDInitiative 1 2012
  • 15. Rodent Audition (Harris): Neural encoding of auditory temporal patterns is highly nonlinear and depends on brain state. clicks clicksInactive Active brain brain state stateTemporal Dynamics of Learning OECDInitiative 1 2012
  • 16. Cortical Representation is Degraded in the Inactivated State Auditory clicks: 400 inactivated 400 activated 400 400 350 350 Rate (Hz) 300 300 Cortex 250 250 200 200 200 200 150 150 100 100 50 50 0 0-1 0 1 2 3 4 5 6 0 0 -1 0 1 2 3 4 5 6 0 1 2 3 4 5 0 1 2 3 4 5 Thalamic Activity Represents the Stimulus in Both States 1000 inactivated 1000 activated 1000 1000 900 900 Thalamus Rate (Hz) 800 800 700 700 600 600 500 500 500 500 400 400 300 300 200 200 100 100 0 0 0 -1 0 1 2 3 4 5 6 0 -1 0 1 2 3 4 5 6 0 1 2 3 4 5 0 1 2 3 4 5 Minces, Harris, Chiba seconds secondsTemporal Dynamics of Learning OECD 2012
  • 17. Burst Spiking of a Single Cortical Neuron Modifies Global Brain StateLi, Poo, and Dan, Science 2009 Temporal Dynamics of Learning Initiative 2 OECD 2012
  • 18. Timescales of investigation Ba   Da   Whisking10s of msec Spikes STDP Phoneme processing Neuromodulation Motor action 100s of msec Perceptual decisions Brain rhythms Action selection Seconds Prediction Temporal Diff. learning Reinforcement Minutes LTP Days Neurogenesis Weeks Spacing effects Years ExpertiseTemporal Dynamics of Learning 18 OECD 2012
  • 19. Gamma Power and Language Learning April Benasich •  Your brain is a giant oscillator (Buzsaki) •  EEG power in the gamma band of this oscillation (~40Hz) has been implicated in attention, perception, memory, language, and neural synchrony •  Gamma power increases throughout development. •  Can we use gamma power in toddlers to predict language ability?Temporal Dynamics of Learning OECD 2012
  • 20. Resting Gamma Power at 36 months of age predictsPredicts Resting Gamma Power In the Frontal Regions language Preschool and 5 yearsScores skills at 4 Language of age From Benasich Lab with HarrisTemporal Dynamics of Learning OECD 2012
  • 21. Brain dynamics matter for learning… •  These results suggests brain dynamics are important for cognitive function •  Early diagnosis means we can now think about early interventions - perhaps a FastForWord for toddlers that can ameliorate the problems that these children face.Temporal Dynamics of Learning OECD 2012
  • 22. • In an inactivated state, cortical processing of the stimulus is degraded in Auditory, Visual, and Tactile modalities. • Thalamic processing of the stimulus is maintained in the inactive state for auditory information but not for visual. • Stimulus fidelity is enhanced by an activated brain state for all modalities. • Cortical activation can be achieved via neuromodulation from a subcortical source (Basal Forebrain Acetylcholine) or locally (a single bursting cortical neuron) or from training. • Natural activation in the form of gamma oscillations in babies predicts later language acquisition. • Early identification can allow for early intervention.Temporal Dynamics of Learning OECD 2012
  • 23. How can we drive temporal processing at rapid timescales, while also drivinggroup synchrony ?Gamelan, Attention, and Synchrony!Temporal Dynamics of Learning OECD 2012
  • 24. Synchronizing with an isochronous beat Adjust following interval Accordingly: 10s of msTarget beat Filter out Non-targets Player Monitor unison Estimate and with target beat: reproduce interval: 10s of ms 100s of msTemporal Dynamics of Learning OECD 2012
  • 25. 10s of ms Details of note shaping, beat to beat alignment 100s of ms Note to note and beat to beat movement 10s of seconds Rhythmic or melodic motives or themes minutes Melodic phrases and chord progressions 10s of minutes - hours Musical structures and compositions Accurate synchrony and integration Amongst players, tooTemporal Dynamics of Learning OECD 2012
  • 26. Research Questions1) Does ability to pay attention correlate with ability to synchronize in a group setting?2) Is this correlation, if it exists, solely due to differences in attentional allocation or do underlying differences in time processing also exist? 3) Since it is possible to improve one s ability to synchronize through group music practice, can such improvements translate into measurable improvements in attentional performance? Temporal Dynamics of Learning OECD 2012
  • 27. The Gamelan Project – preliminary study Does ability to pay attention correlate with ability to synchronize in a group setting?Temporal Dynamics of Learning OECD 2012
  • 28. Experimental instrument: aluminum keys with piezoelectric film elementsTemporal Dynamics of Learning OECD 2012
  • 29. Audio recording data Temporal Dynamics of Learning OECD 2012
  • 30. Examples of good and poor synchronizers using vector strength analysis 3 3 2 2 1 1 phase 0 poor 0 -1 -1 -2 -2 -3 -3 0 10 20 30 40 50 60 14 12 10 8 6 4 2 0 3 3 2 2 1 1 0 phase 0 good -1 -1 -2 -2 -3 -3 0 10 20 30 40 50 60 14 12 10 8 6 4 2 0Temporal Dynamics of Learning OECD 2012
  • 31. Other measures Swann teacher ratings Psychometric testsResults A significant correlation was found across all measures p<0.0025Temporal Dynamics of Learning OECD 2012
  • 32. Temporal Dynamics of Learning OECD 2012
  • 33. Temporal Dynamics of Learning OECD 2012
  • 34. Passive Listening. Sounds are all the same. We only care about timing. deviantTemporal Dynamics of Learning OECD 2012
  • 35. Example A Short Condition Long Condition Voltage uV standard deviantTemporal Dynamics of Learning Time OECD 2012
  • 36. Example A Short Condition Long Condition Voltage uV Sensitivity to Time standard deviantTemporal Dynamics of Learning Time OECD 2012
  • 37. Synchronization Subject drums to a steady beat, even when sounds are omitted.Temporal Dynamics of Learning OECD 2012
  • 38. Drumming Data AccuracyTemporal Dynamics of Learning OECD 2012
  • 39. sound omitted Time (seconds)Temporal Dynamics of Learning OECD 2012
  • 40. Conclusions: •  There exists a correlation between the ability to synchronize and classic measures of attention •  We have developed measures of individual differences in temporal processing at the behavioral and neural level •  These differences correlate with other cognitive characteristics across individuals •  A tool to measure the effects of musical training on measures of attention and classroom performanceTemporal Dynamics of Learning OECD 2012
  • 41. Neuroimaging techniques provide a window into The structural development necessary for timing Courtesy of Terry JerniganTemporal Dynamics of Learning OECD 2012
  • 42. How does white matter maturation relate to readiness for academic skills? Can development be facilitated?Thus far, longitudinal DTI findings support continued microstructural change in white matter during late adolescence, and suggest ongoing refinement ofprojection and association fibers into early adulthood. Structural asymmetries are correlated with measures several measures of academic performance. Temporal Dynamics of Learning OECD 2012
  • 43. Why does time matter?•  Time matters for processing•  Time matters for learning•  Time matters for action•  Thus, optimizing temporal processing is a key feature of the ability to learn effectivelyTemporal Dynamics of Learning OECD 2012
  • 44. Timescales of investigation Ba   Da   Whisking10s of msec Spikes STDP Phoneme processing Neuromodulation Motor action 100s of msec Perceptual decisions Brain rhythms Action selection Seconds Prediction Temporal Diff. learning Reinforcement Minutes LTP Days Neurogenesis Weeks Spacing effects Years ExpertiseTemporal Dynamics of Learning OECD 2012
  • 45. What are the implications for education? Timing is an essential property of the world and the brain and it must be optimal for optimal learning to occur. Plasticity is an inherent feature of the brain that can be optimized and remediated through specific types of behavioral training. Basic aspects of our culture including music, exercise, and enrichment may facilitate learning and cognition. We should be circumspect before removing them from the curriculum.Temporal Dynamics of Learning OECD 2012