Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Age-related and individual differences in the time-course and information content of early visual brain activity

Age-related and individual differences in the time-course and information content of early visual brain activity

In this talk I will present a quantification of age-related and individual differences in visual processing. I will show examples using simplified pictures of faces and textures, EEG recordings, and relatively simple tasks, demonstrating that ageing, at least in a cross-sectional design involving 120 participants, is associated with a slowing-down of visual processing. This ageing effect does not seem to be explainable by low-level factors, such as retinal illuminance, and these factors also fail to explain the large individual differences in processing speed. Using reverse-correlation and mutual information, I will describe how we can infer the information content of early face ERPs. These face ERPs are mostly modulated by the presence of the contralateral eye area in both younger and older participants. However, this contralateral eye sensitivity is delayed and weaker in older adults. Finally, recent experiments address the stimulus and task-specificity of age-related ERP differences.

Guillaume Rousselet

March 08, 2017
Tweet

Other Decks in Research

Transcript

  1. Age-related and individual differences in the time-course and information content

    of early visual brain activity Guillaume Rousselet Magdalena Bieniek, Kasia Jaworska, Hannah Gilman Robin Ince, Nicola Van Rijsbergen Philippe Schyns 1 Cyril Pernet Allison Sekuler Patrick Bennett @robustgar https://garstats.wordpress.com
  2. Motulsky, H.J. (2014) Common misconceptions about data analysis and statistics.

    J Pharmacol Exp Ther, 351, 200-205. https://garstats.wordpress.com/2016/05/27/the-percentile-bootstrap/
  3. Rousselet, Foxe & Bolam (2016) A few simple steps to

    improve the description of group results in neuroscience
  4. Age-related differences in visual processing speed Rousselet, Husk, Pernet, Gaspar,

    Bennett & Sekuler, BMC Neuroscience 2009 Rousselet, Gaspar, Pernet, Husk, Bennett & Sekuler, Front. Psychology 2010 Photo Credit JD Howell
  5. 8 @robustgar https://garstats.wordpress.com Is perception getting slower as we get

    older? In a biased cross-sectional sample of older adults recruited in Hamilton (Ontario, Canada) & Glasgow, can we find evidence for delayed brain activity, as measured with scalp EEG, in response to overly simplified visual stimuli?
  6. Tracking ERP sensitivity to phase noise  using a parametric

    design task stimulus manipulation Rousselet, Pernet, Bennett & Sekuler, BMC Neuroscience 2008 Rousselet, Husk, Pernet, Gaspar, Bennett & Sekuler, BMC Neuroscience 2009 Rousselet, Gaspar, Pernet, Husk, Bennett & Sekuler, Front. Psychology 2010 Focus on accuracy, not speed
  7. ERPs to face & noise images ~ age face trials

    noise trials 0.5 0 -0.5 0.5 0 -0.5 ERP amplitude (µV/cm2) ERP amplitude (µV/cm2) YOUNG OLD LAT: 159 ms [156, 167] AMP: -0.62 µV/cm2 [-0.77, -0.45] LAT: 141 ms [136, 146] AMP: -0.52 µV/cm2 [-0.70, -0.40] P1 N170 LAT: 152 ms [141, 161] AMP: 0.01 µV/cm2 [-0.15, 0.10] LAT: 175 ms [164, 185] AMP: -0.47 µV/cm2 [-0.64, -0.31] A face - noise 0.5 0 -0.5 ERP amplitude (µV/cm2) - -200 0 200 400 600 800 1000 -100 100 300 500 700 900 -200 0 200 400 600 800 1000 -100 100 300 500 700 900 -200 0 200 400 600 800 1000 -100 100 300 500 700 900 face trials noise trials 0.5 0 -0.5 0.5 0 -0.5 ERP amplitude (µV/cm2) ERP amplitude (µV/cm2) YOUNG OLD LAT: 159 ms [156, 167] AMP: -0.62 µV/cm2 [-0.77, -0.45] LAT: 141 ms [136, 146] AMP: -0.52 µV/cm2 [-0.70, -0.40] P1 N170 LAT: 152 ms [141, 161] AMP: 0.01 µV/cm2 [-0.15, 0.10] LAT: 175 ms [164, 185] AMP: -0.47 µV/cm2 [-0.64, -0.31] A face - noise 0.5 0 -0.5 ERP amplitude (µV/cm2) - -200 0 200 400 600 800 1000 -100 100 300 500 700 900 -200 0 200 400 600 800 1000 -100 100 300 500 700 900 -200 0 200 400 600 800 1000 -100 100 300 500 700 900 face trials noise trials 0.5 0 -0.5 0.5 0 -0.5 ERP amplitude (µV/cm2) ERP amplitude (µV/cm2) YOUNG OLD LAT: 159 ms [156, 167] AMP: -0.62 µV/cm2 [-0.77, -0.45] LAT: 141 ms [136, 146] AMP: -0.52 µV/cm2 [-0.70, -0.40] P1 N170 LAT: 152 ms [141, 161] AMP: 0.01 µV/cm2 [-0.15, 0.10] LAT: 175 ms [164, 185] AMP: -0.47 µV/cm2 [-0.64, -0.31] A face - noise 0.5 0 -0.5 ERP amplitude (µV/cm2) - -200 0 200 400 600 800 1000 -100 100 300 500 700 900 -200 0 200 400 600 800 1000 -100 100 300 500 700 900 -200 0 200 400 600 800 1000 -100 100 300 500 700 900 Electrode locations. of electrodes included in three subsets: midline (CE, red), posterior-lateral in the left ere (LE, green) and in the right hemisphere (RE, blue). LE CE RE
  8. Rousselet, Gaspar, Pernet, Husk, Bennett & Sekuler, Front. Psychology 2010

    Bieniek, Frei & Rousselet, Front. Psychology 2013 Normalised t2 Age quantiles in years Time in ms Age-related differences in t² time-courses
  9. Age-related differences in t² time-courses Rousselet, Gaspar, Pernet, Husk, Bennett

    & Sekuler, Front. Psychology 2010 Bieniek, Frei & Rousselet, Front. Psychology 2013 Time in ms Weighted mean of t2 functions 0 50 100 150 200 250 300 350 400 450 500 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 21 22 23 26 30 33 37 41 45 50 56 61 63 65 67 68 70 Normalised data Time in ms Weighted mean of t2 functions 0 50 100 150 200 250 300 350 400 450 500 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 21 22 23 26 30 33 37 41 45 50 56 61 63 65 67 68 70 Normalised cumulated data age difference starts ~ 120 ms
  10. slope = 1 ms / year Rousselet, Gaspar, Pernet, Husk,

    Bennett & Sekuler, Front. Psychology 2010 Bieniek, Frei & Rousselet, Front. Psychology 2013 Age-related delay in face sensitivity
  11. Other age-related differences? • Face / noise ERP onsets •

    Effect sizes • Lateralization • Within session variability • Between session variability • Task modulations ✘ ✘ ✘ ✘ ✘ ✘ Bieniek et al. EJN 2015
  12. Low-level confound: senile miosis Bieniek, Frei & Rousselet, Front. Psychol.

    2013 Winn et al. Inv. Ophthal. & Vis. Sci. 1994 59 participants (18-79), 51 tested twice
  13. Can we match the ERPs of young participants at low

    luminance with the ERPs of old participants at high luminance? Bieniek, Frei & Rousselet, Front. Psychol. 2013
  14. Pupil size: age & luminance 10 20 30 40 50

    60 70 80 100 150 200 250 300 350 400 Age 50% integration time (ms) Bright to dark 10 100 150 200 250 300 350 400 0 1 2 slope 50IT / age 300 tr IT 0 1 2 300 slope = 1 ms / year Processing speed: age & luminance
  15. Pupil size: individual speed differences? 10 20 30 40 50

    60 70 80 100 150 200 250 300 350 400 Age 50% integration time (ms) Bright to dark 10 20 30 40 50 60 70 80 100 150 200 250 300 350 400 Age Dark to bright 0 1 2 slope 50IT / age 300 r IT 0 1 2 300 Processing speed: age & luminance −3 −2 −1 0 1 2 −80 −60 −40 −20 0 20 40 60 80 100 Residuals Pupil / Age Residuals 50% IT / Age −3 −80 −60 −40 −20 0 20 40 60 80 100 0 1 slope 50IT / 0 10 60.8 31 16 8.16 4.19 2.17 1.12 0.59 60.8 100 200 300 intr 50IT 0 1 0 10 100 200 300 ope IT/age / upil/age Bieniek, Frei & Rousselet, Front. Psychology 2013
  16. Can we make young adults’ brain activity look old by

    changing their pupil size? Magda Bieniek Bieniek, Frei & Rousselet, Front. Psychol. 2013
  17. Example: Sampling Stimulus Trials Bubble Mask = Detection task response

    • Face • Noise Face or Noise Image × slide from Robin Ince
  18. Behaviour Presented Stimulus Bubble Mask = Presented Stimulus Bubble Mask

    = × Presented Stimulus Bubble Mask = ge × • Which pixels affect reaction time? • Quantify this with Mutual Information (MI) Pixel visibility Reaction time Rousselet et al. Journal of Vision (2014) slide from Robin Ince
  19. Mutual information(face pixels, RT) Rousselet, Ince, van Rijsbergen & Schyns,

    J. Vis. 2014 Mutual information code + tutorial: Ince et al. Hum. Brain Mapp. 2016
  20. MI(face pixels, ERP) Jaworska et al. (in preparation) Figure 5

    Time-courses of the maximum MI across pixels.
  21. ERPs to face & noise images ~ age face trials

    noise trials 0.5 0 -0.5 0.5 0 -0.5 ERP amplitude (µV/cm2) ERP amplitude (µV/cm2) YOUNG OLD LAT: 159 ms [156, 167] AMP: -0.62 µV/cm2 [-0.77, -0.45] LAT: 141 ms [136, 146] AMP: -0.52 µV/cm2 [-0.70, -0.40] P1 N170 LAT: 152 ms [141, 161] AMP: 0.01 µV/cm2 [-0.15, 0.10] LAT: 175 ms [164, 185] AMP: -0.47 µV/cm2 [-0.64, -0.31] A face - noise 0.5 0 -0.5 ERP amplitude (µV/cm2) - -200 0 200 400 600 800 1000 -100 100 300 500 700 900 -200 0 200 400 600 800 1000 -100 100 300 500 700 900 -200 0 200 400 600 800 1000 -100 100 300 500 700 900 face trials noise trials 0.5 0 -0.5 0.5 0 -0.5 ERP amplitude (µV/cm2) ERP amplitude (µV/cm2) YOUNG OLD LAT: 159 ms [156, 167] AMP: -0.62 µV/cm2 [-0.77, -0.45] LAT: 141 ms [136, 146] AMP: -0.52 µV/cm2 [-0.70, -0.40] P1 N170 LAT: 152 ms [141, 161] AMP: 0.01 µV/cm2 [-0.15, 0.10] LAT: 175 ms [164, 185] AMP: -0.47 µV/cm2 [-0.64, -0.31] A face - noise 0.5 0 -0.5 ERP amplitude (µV/cm2) - -200 0 200 400 600 800 1000 -100 100 300 500 700 900 -200 0 200 400 600 800 1000 -100 100 300 500 700 900 -200 0 200 400 600 800 1000 -100 100 300 500 700 900 face trials noise trials 0.5 0 -0.5 0.5 0 -0.5 ERP amplitude (µV/cm2) ERP amplitude (µV/cm2) YOUNG OLD LAT: 159 ms [156, 167] AMP: -0.62 µV/cm2 [-0.77, -0.45] LAT: 141 ms [136, 146] AMP: -0.52 µV/cm2 [-0.70, -0.40] P1 N170 LAT: 152 ms [141, 161] AMP: 0.01 µV/cm2 [-0.15, 0.10] LAT: 175 ms [164, 185] AMP: -0.47 µV/cm2 [-0.64, -0.31] A face - noise 0.5 0 -0.5 ERP amplitude (µV/cm2) - -200 0 200 400 600 800 1000 -100 100 300 500 700 900 -200 0 200 400 600 800 1000 -100 100 300 500 700 900 -200 0 200 400 600 800 1000 -100 100 300 500 700 900 Jaworska et al. (in preparation)
  22. Age-related delays: stimulus specificity? N170 to noise: top-down effects? Figure

    18 Stimuli. Top row shows all ten identities of faces, the second row shows all ten images of houses, and the n=24 younger (20-39, median = 22) n=24 older (59-85, median = 67.5) Jaworska et al. (in preparation)
  23. Age-related delays: stimulus specificity? 0 0.1 0.2 0.3 0.4 0.5

    MI (bits) Face vs noise young older -0.2 -0.1 0 0.1 0.2 0.3 -0.2 -0.1 0 0.1 0.2 0.3 0 0.1 0.2 0.3 0.4 0.5 0 0.1 0.2 0.3 0.4 0.5 House vs noise Letter vs noise -100 0 100 200 300 400 500 -100 0 100 200 300 400 500 -100 0 100 200 300 400 500 -100 0 100 200 300 400 500 -100 0 100 200 300 400 500 -100 0 100 200 300 400 500 -0.2 -0.1 0 0.1 0.2 0.3 MI (bits) Time (ms) Time (ms) Time (ms) A 0.8 max MI 250 300 MI latency 50% IT s) ) B Time in ms Weighted mean of t2 functions 0 50 100 150 200 250 300 350 400 450 500 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 21 22 23 26 30 33 37 41 45 50 56 61 63 65 67 68 70 Jaworska et al. (in preparation)
  24. Is perception getting slower as we get older? Alteration of

    information content? Behavioural consequences? Domain specificity? Neuronal causes: myelin, GABA, network properties…?