Stimulus Trial 1 1 0 ms 1800ms Trial 2 Oms 1800ms Trial 3.. 0t 1800ms Trial i I Oms 1800ms RIDE Find latency-variable components Locate the latency in single trials Reconstruct latency-corrected ERP Improve understanding of brain-behavior relationship UllIS ⊥。 VEILS 1/28/2021
Stimulus . . . 0ms 1800ms 0ms 1800ms 0ms 1800ms 0ms 1800ms 0ms 1800ms Trial 1 Trial 2 Trial 3 Trial i Average ERP Reconstructed ERP RIDE ➢ Find latency-variable components ➢ Locate the latency in single trials ➢ Reconstruct latency-corrected ERP ➢ Improve understanding of brain-behavior relationship 1/28/2021 11
Alternative model of ERP: temporal superposition EG()=C1t-1)+C2t-t2)+…+Cn(t-t3n)+5(t) signal →> outpu signal output signal output Mathematically they are the same Previous attempts (with all time markers known) X Cannot deal with latency-unknown components X Suffer from strong low-frequency distortion Fourier decomposition Hansen, 1983 (low-frequency divergence) Takeda et al, 2008 (no asymmetry in R) k=1 Iterative de-convolution Wordorff. 1993 Zhang. 1998 General Linear Model decomposition (least square-based Dandekar. et al. 2012 1/28/2021 agency jitter
𝐸𝐸𝐺𝑖 𝑡 = 𝐶1 𝑡 − 𝜏1𝑖 + 𝐶2 𝑡 − 𝜏2𝑖 + ⋯ + 𝐶𝑛 𝑡 − 𝜏3𝑛 + 𝜉𝑖 (𝑡) A A A B B B C C C signal signal signal output output output … … … Alternative model of ERP: temporal superposition Previous attempts: (with all time markers known) —Fourier decomposition Hansen, 1983 (low-frequency divergence) Takeda et al, 2008 (no asymmetry in R) —Iterative de-convolution Wordorff, 1993 Zhang, 1998 — General Linear Model decomposition (least square-based) Dandekar, et al, 2012 Mathematically they are the same! ✘ Cannot deal with latency-unknown components ✘ Suffer from strong low-frequency distortion 0 10 20 30 40 50 0 10 20 30 40 50 60 latency jitter error stength k=1 k=2 k=3 k=4 k=5 1/28/2021 12
How does the single trials look like 800 1000 400 800 time after stimulus(ms) Response locked component cluster R Stimulus locked Central component component cluster S cluster c This is a single trial ERP data synchronized to stimulus onset and sorted by reaction time from a speech production data (from Or 1/28/2021 This is a very excellent data showing dissociated component
1/28/2021 13 How does the single trials look like - This is a single trial ERP data synchronized to stimulus onset and sorted by reaction time from a speech production data (from Oz). - This is a very excellent data showing dissociated component clusters, but what if the components are serious overlapped? Stimulus locked component cluster S Response locked component cluster R Central component cluster C
How does the single trials look like 600 800 1000 0040060080010001200 time after stimulus(ms) Response locked component cluster R Stimulus locked Central component component cluster S cluster c This is a single trial ERP data synchronized to stimulus onset and sorted by reaction time from a speech production data (from Oz) 1/28/2021
1/28/2021 14 How does the single trials look like - This is a single trial ERP data synchronized to stimulus onset and sorted by reaction time from a speech production data (from Oz). Stimulus locked component cluster S Response locked component cluster R Central component cluster C ? ? ?
RIDE, basic framwork 1. Initial estimation of latency of c :2. Decomposition Module 3. Re-estimation of latency of C Obtain S, C and R based on Ls, Lc and LR Remove s and r from single trials and timate the latency of c by template matching between C and residue Convergence 1/28/2021
1/28/2021 15 RIDE: basic framwork 3. Re-estimation of latency of C Remove S and R from single trials and estimate the latency of C by template matching between C and residue. 1. Initial estimation of latency of C 2. Decomposition Module Obtain S, C and R based on Ls , Lc and LR convergence