残差迭代分解 Residue Iteration Decomposition（RIDE）Restoring latency-variable ERP components from single trials

Sometimes we respond slow, sometimes fast Residue Iteration Decomposition(RIDE) Restoring latency-variable ERP components from single trials Intra-person inter-person variability: y Difficulty for cognitive data analysis subject 2 f Opportunity for studying brain activity-function relation subiect 3 RIDE: ☆Un- mix ERP components 0001200 40 Restore erp component shapes React on time(ms) o Get useful information about sub-process variabl ity in single trials Team. Guang ouyan Werner sommer Changsong Zhou a Department of psychology sh: Dept. Physics, Center for Nonlinear Studies, 3 2 Humboldt University Berlin, Germany Hong Kong Baptist University BER1少 1/28/2021

Residue Iteration Decomposition (RIDE) Restoring latency-variable ERP components from single trials Sometimes we respond slow, sometimes fast Guang Ouyang Changsong Zhou Dept. Physics, Center for Nonlinear Studies, ICTS Hong Kong Baptist University Werner Sommer Department of Psychology Humboldt University Berlin, Germany Team: 1/28/2021 1 Intra-person & inter-person Variability:  Difficulty for cognitive data analysis  Opportunity for studying brain activity-function relation RIDE: ❖ Un-mix ERP components ❖ Restore ERP component shapes ❖ Get useful information about sub-process variability in single trials

Outline ☆ Motivation: Problems in erP due to latency variability Alternative model for erp Latency variable components ☆R|DE Component decomposition ERP reconstruction Single trial variability Comparison with ICA Applications of ride in neuropsychology ☆ Summary 1/28/2021

Outline ❖Motivation: Problems in ERP due to latency variability ❖Alternative model for ERP: Latency variable components ❖RIDE ➢ Component decomposition ➢ ERP reconstruction ➢ Single trial variability ➢ Comparison with ICA ➢ Applications of RIDE in neuropsychology ❖Summary 1/28/2021 2

How do ERP researchers deal with "noisy"EEG signal? EEG ERP average 5 00 Time(ms) Time(ms) Stimulus onset ERP (Event-related Potential) 1/28/2021

How do ERP researchers deal with “noisy” EEG signal? EEG ERP -200 0 200 400 600 800 1000 Time (ms) -200 0 200 400 600 8001000 -15 -10 -5 0 5 10 15 20 25 Votage (V) Time (ms) (a) (b) average 50 μ V Stimulus onset ERP (Event-related Potential) 1/28/2021 3

Averaging ERPs: Basic Idea EEG following: evoked signal◆ random noi· stimulus MET.24 Random Noise 1/N MM w- NW Constant Signal Averaging Average E P Signal Average noise Number of Averagings (N) 1/28/2021

Averaging ERPs: Basic Idea 1/28/2021 4

ERP components as probes into cognitive stages Stimulus Stage 1 Stage 2 Stage n Response ERP-component 1/28/2021

ERP components as probes into cognitive stages Stage 1 Stage 2 Stage n Stimulus Response ERP-component C1 C2 Cn 1/28/2021 5

Assumption of averaging ERP implying. gn sga→DA→[B signal A 3 ubject 1 signal A subject 3 2 jga→[A」 B sign B→ 0 40060m0100012001400 Oms 200ms 400ms Reaction time(ms) Identical processes in each realization Contradictory to strong response variability 1/28/2021

A A A A A A B B B B B B C C C C C C signal signal signal signal signal signal output output output output output output … … … … … … Identical processes in each realization; Contradictory to strong response variability Assumption of averaging ERP implying… 0ms 200ms 400ms 600ms 800ms 1000ms 1/28/2021 6

Alternative assumption about brain response signal output signal 1=A B C output sgna→[A output gna >A sIg output signal一 A→>B output signal A output Oms 200ms 400ms 600ms 800ms 1000ms Latency jitters in each realization 1/28/2021

Alternative assumption about brain response A A A A A A B B B B B B C C C C C C signal signal signal signal signal signal output output output output output output … … … … … … 0ms 200ms 400ms 600ms 800ms 1000ms Latency jitters in each realization 1/28/2021 7

Reality: systematic variability in the EEG data stimulus onset reaction Times Sorting by rT stimulus reaction time WAAM 900ms Electroencephalographic(EEG) What is the consequence of latency variability on ERP? 1/28/2021 8

Reality: systematic variability in the EEG data time trial index 100 300 500 700 900 ms 20 40 60 80 100 120 140 160 stimulus onset reaction Times 1/28/2021 8 What is the consequence of latency variability on ERP? Sorting by RT

atency variability: Smearing effect (convolution s(t) TXXXXXWX s(t*p(t) Generally the convolution reduces the amplitude When the spreading'distribution is broad the convolution approaches to zero 1/28/2021

Latency variability: Smearing effect (convolution) 𝑆 𝑡 𝑆 𝑡 ∗ 𝜌(𝑡) ➢Generally, the convolution reduces the amplitude ➢When the ‘spreading’ distribution is broad, the convolution approaches to zero 1/28/2021 9

Stimulus Trial 1 1 0 ms 1800ms Trial 2 Oms 1800ms Limitations of average ERP method Smearing and mixing of components Limited precision of amplitude/ latency and conditional effects Vague interpretation in mental chronometry Lose valuable single trial dynamic information A possible solution? Average erp Oms 1800ms 1/28/2021

Stimulus . . . 0ms 1800ms 0ms 1800ms 0ms 1800ms 0ms 1800ms 0ms 1800ms Trial 1 Trial 2 Trial 3 Trial i Average ERP Limitations of average ERP method ➢ Smearing and mixing of components ➢ Limited precision of amplitude/latency and conditional effects ➢ Vague interpretation in mental chronometry ➢ Lose valuable single trial dynamic information A possible solution? 1/28/2021 10