neuRosim Outline Motivation Simulating fMRI data: the R package The need for simulation The need for neuRosim validated simulation The need for software Features Goals What can you do Marijke Welvaert with neuRosim? How is neuRosim organized? Example Department of Data Analysis Setting up the design Ghent University Simulating the data Exporting and analyzing the data Berlin Workshop on Statistics and Neuroimaging 2011 Summary
Outline neuRosim 1 Motivation The need for simulation Outline The need for validated simulation Motivation The need for The need for software simulation The need for validated simulation 2 Features The need for software Features Goals Goals What can you do with neuRosim? What can you do with neuRosim? How is neuRosim organized? How is neuRosim organized? Example 3 Example Setting up the design Setting up the design Simulating the data Exporting and Simulating the data analyzing the data Summary Exporting and analyzing the data 4 Summary
Knowing the ground truth in MRI neuRosim Outline Motivation The need for simulation The need for validated simulation The need for software Features Goals What can you do with neuRosim? How is neuRosim organized? Example Setting up the design Simulating the data Exporting and analyzing the data Summary
Knowing the ground truth in fMRI neuRosim Outline Motivation The need for simulation The need for validated simulation The need for software Features Goals What can you do with neuRosim? How is neuRosim organized? Example Setting up the design Simulating the data Exporting and analyzing the data Summary
Reflection in the literature neuRosim Web of Science publications/year Outline Motivation 3000 The need for simulation The need for validated simulation 2500 The need for software Number of publications Features 2000 Goals What can you do with neuRosim? 1500 How is neuRosim organized? Example 1000 Setting up the design Simulating the data 500 Exporting and analyzing the data Summary 0 1995 2000 2005 2010 Year of publication
fMRI data components neuRosim Activation experimentally induced Outline Motivation spontaneous The need for simulation Known artefacts The need for validated simulation B 0 inhomogeneities The need for software Features low-frequency drift Goals What can you do Noise with neuRosim? How is neuRosim system organized? Example movement Setting up the design Simulating the data physiological Exporting and analyzing the data task-related Summary . . . Spatial and temporal correlations
Typical fMRI simulation studies neuRosim 1 hybrid simulation Outline Motivation known activation combined with real noise The need for e.g. Bianciardi et al. (2004), Lange (1999), Weibull et al. (2008) simulation The need for 2 white time series validated simulation The need for software known activation combined with white noise Features i.i.d or AR(1) Gaussian distribution Goals What can you do e.g. Lei et al. (2010), Lin et al. (2010), Purdon & Weisskoff with neuRosim? How is neuRosim (1998), Smith et al. (2011) organized? 3 other Example Setting up the design model-based simulation, Bloch equations, noise based on Simulating the data Exporting and residuals of real data analyzing the data e.g. Drobnjak et al. (2006), Havlicek et al. (2010), Logan & Summary Rowe (2004), Ramsey et al. (2010)
Problems - Discrepancies - Shortcomings neuRosim Outline Motivation real noise may contain undesired activity The need for simulation The need for simulated noise = system noise validated simulation The need for software beware of the phrase: Features “. . . simulations under realistic noise conditions. . . ” Goals What can you do total ignorance of spatial context with neuRosim? How is neuRosim organized? no stand-alone simulations Example often missing (crucial) information while reporting simulation Setting up the design Simulating the data studies Exporting and analyzing the data Summary
The choice of simulation model matters! neuRosim Outline Low CNR High CNR Motivation 0.35 1.00 The need for simulation The need for 0.95 0.30 validated simulation The need for software 0.90 0.25 Features 0.85 Goals Power Power 0.20 What can you do with neuRosim? 0.80 How is neuRosim 0.15 organized? 0.75 Example 0.10 0.70 White noise Setting up the design AR(1) noise Phys. noise Simulating the data Real noise 0.05 0.65 Exporting and analyzing the data 0.02 0.04 0.06 0.08 0.10 0.20 0.25 0.30 0.35 Summary Contrast−to−noise ratio Contrast−to−noise ratio
Towards a convergence of simulation methods neuRosim Outline in-house developed software routines, often not available for the Motivation community The need for simulation language barrier The need for validated simulation The need for software no widespread software packages Features Goals What can you do with neuRosim? But. . . How is neuRosim organized? Example POSSUM (FSL) Setting up the design Simulating the data DCM simulator (SPM) Exporting and analyzing the data simtb (Matlab Toolbox) Summary
neuRosim wants neuRosim Outline Motivation The need for simulation The need for validated simulation to provide a tool for simulating fMRI data The need for software to be a base for more validated simulation studies Features Goals to make simulation available for less technical researchers What can you do with neuRosim? How is neuRosim to allow maximum flexibility for the useRs organized? Example Setting up the design Simulating the data Exporting and analyzing the data Summary
What can you do with neuRosim? neuRosim Outline Motivation The need for specify your experimental design based on stimulus onsets and simulation The need for durations validated simulation The need for software specify activated regions using an xyz-coordinate system Features Goals simulate BOLD activation with the choice of different models What can you do with neuRosim? simulate resting state activation (still under development) How is neuRosim organized? simulate fMRI noise originating from different noise sources Example Setting up the design generate fMRI data from 1D time series to 4D volume data Simulating the data Exporting and analyzing the data Summary
Low-level functions neuRosim Outline Building blocks for advanced useRs who want in-depth control over Motivation their simulation data The need for simulation The need for validated simulation Activation functions Noise functions The need for software Features Goals stimfunction() systemnoise() What can you do with neuRosim? specifydesign() temporalnoise() How is neuRosim organized? specifyregion() spatialnoise() Example lowfreqdrift() Setting up the design canonicalHRF() physnoise() Simulating the data gammaHRF() Exporting and tasknoise() analyzing the data balloon() Summary
High-level functions neuRosim Outline Motivation The need for Direct simulation of fMRI data simulation The need for validated simulation The need for software Preparation functions Simulation functions Features Goals What can you do simTSfmri() with neuRosim? simprepTemporal() How is neuRosim simVOLfmri() organized? simprepSpatial() Example simRestingStatefmri() Setting up the design Simulating the data Exporting and analyzing the data Summary
Real dataset neuRosim Consider the data from a repetition priming experiment performed using event-related fMRI (Henson et al. , 2002). Outline Motivation 2 × 2 factorial design The need for simulation famous vs non-famous faces The need for validated simulation effect of repetition The need for software Features Goals What can you do with neuRosim? How is neuRosim organized? Example Setting up the design Simulating the data Exporting and analyzing the data Summary
Setting up the design (1) neuRosim Temporal Parameters Outline R> nscan <- 351 Motivation R> TR <- 2 The need for R> total.time <- nscan*TR simulation R> onsets.N1 <- c( 6.75, 15.75, 18.00, 27.00, 29.25, 31.50, The need for + 36.00, 42.75, 65.25, 74.25, 92.25, 112.50, 119.25, validated simulation The need for software + 123.75, 126.00, 137.25, 141.75, 144.00, 146.25, 155.25, + 159.75, 162.00, 164.25, 204.75, 238.50)*TR Features R> onsets.N2 <- c(13.50, 40.50, 47.25, 56.25, 90.00, 94.50, Goals + 96.75, 135.00, 148.50, 184.50, 191.25, 202.50, 216.00, What can you do with neuRosim? + 234.00, 236.25, 256.50, 261.00, 281.25, 290.25, 303.75, How is neuRosim + 310.50, 319.50, 339.75, 342.00)*TR organized? R> onsets.F1 <- c( 0.00, 2.25, 9.00, 11.25, 22.50, 45.00, Example + 51.75, 60.75, 63.00, 76.50, 78.75, 85.50, 99.00, Setting up the design + 101.25, 103.50, 117.00, 130.50, 150.75, 171.00, 189.00, Simulating the data + 227.25, 265.50, 283.50, 285.75, 288.00, 344.25)*TR Exporting and R> onsets.F2 <- c(33.75, 49.50, 105.75, 153.00, 157.50, 168.75, analyzing the data + 177.75, 180.00, 182.25, 198.00, 222.75, 240.75, 254.25, Summary + 267.75, 270.00, 274.40, 294.75, 299.25, 301.50, 315.00, + 317.25, 326.25, 333.00, 335.25, 337.50, 346.50)*TR R> onsets <- list(onsets.N1, onsets.N2, onsets.F1, onsets.F2) R> dur <- list(0, 0, 0, 0)
Recommend
More recommend
