REM Sleep scupts memories

• REM sleep

  Rapid eye movement sleep
• Michelangelo: 'The sculpture is already complete within the marble block, before I start my work. It is already there, I just have to chisel away the superfluous material.'
• REM sleep

  Sculpts our memories
• Does sleep help memory by chiseling away the superfluous material?
• Why care about REM sleep?

  • Learning/Memory consolidation
  • Rescue of weak memories
  • Dreams and emotional memories
  • Are there REM EEG events related to behavior?
  • Can we develop hypothesis of REM's function?
• 90 minute Sleep Cycle

  • 1-2% of sleep
  • Myoclonic jerks
  • Hypnogogic dreams
  • 60% of sleep
  • Sleep Spindles
  • 20% of sleep
  • Slow wave activity
  • Delta
  • 20% of sleep
  • Rapid eye movement
  • Atonia
  • Alpha, Theta
• NIGHTTIME sleep

  • 11PM
  • 12AM
  • 1AM
  • 2AM
  • 3AM
  • 4AM
  • 5AM
  • 6AM
• DAYTIME sleep

  • 7PM
  • 5PM
  • 3PM
  • 1PM
  • 11AM
  • 9AM
• Experimental Timeline

  1. 60' NAP

  2. No-NAP

  3. 90' NAP vs NREM

  4. NREM + REM

  5. 9:00

  6. 12:00

  7. 13:00

  8. 15:00

  9. 16:00

  10. 19:00
• REM-dependent Perceptual Learning

  Decreasing Performance
• Improvement correlated with SWS X REM


  r = .37, p=.01
• Nap is as Good as a Night


  Improvement in performance
• Perception is vulnerable to perceptual fatigue
• Perceptual Learning is optimized (i.e., consolidated) during offline Sleep, compared with Wake
• Distinct roles for sleep stages

  • NREM: Protects memories from perceptual fatigue
  • REM: Enhances perceptual learning
• Neuro-milieu of NREM and REM

  • NREM: GABA
  • REM: ACh
• Brain states from offline to online


  • Active Wake
  • Quiet Wake
  • Sleep
• NREM


  Plasticity +, Sensory Input +
• REM

  Plasticity +, Sensory Input 0
• Quiet Wake

  Plasticity 0, Sensory Input 0
• Active Wake

  Plasticity 0, Sensory Input +
• Does memory function similarly across all four brain states?
• Does the strength of encoding (e.g., weak memories vs strong memories) play a role in consolidation?
• Interference Paradigm

  • No interference
  • Proactive
  • Retroactive
• Interference decreased memory strength

  Main Effect of Interference
• Performance Profiles

  • No Interference
  • Retroactive
  • Proactive
• SWS and REM

  Predict learning improvement
• Opportunistic Consolidation Hypothesis
• Dreaming about memories

  Maze related vs Not maze related
• Dreaming increases Emotional Memory Trade-off

  Only Dreamers showed emotional memory tradeoff
• Dreaming reduces reactivity

  Only Dreamers had decreased valence and arousal ratings for negative images overnight
• Dream mood affects reactivity

  Positive dreams led to decreased valence rating overnight to neutral, but not negative images
• Dreaming actively modulates emotional memory processing
• Dreaming may be critical for active forgetting
• Can we track dream timing, content, emotional details, in real-time?
• Does sleep stage matter?
• What are the mechanisms, EEG signatures, of forgetting during REM?
• Characterize REM burst events

  1. Visual examination of EEG

  2. Design an algorithm to detect EEG burst events during REM sleep

  3. Examine bursts in relation to learning and memory
• Theta burst and perceptual learning

  Retinotopically specific increases in REM theta bursts in trained visual field, not associated with consolidation
• Alpha burst and verbal memory


  REM Alpha bursts associated with greater encoding and more sleep-dependent forgetting of episodic memories