Part II - Wellcome Trust Centre for Neuroimaging

Part II - Wellcome Trust Centre for Neuroimaging

Kristy DeDuck & Luzia Troebinger MFD Wednesday 18th January 2012 Image time-series Realignment Spatial filter Design matrix

Smoothing General Linear Model Normalisation Anatomical reference Parameter estimates

Statistical Parametric Map Statistical Inference RFT p <0.05 Overview

Experimental Design Types of Experimental Design Timing parameters Blocked and Event-Related & Mixed design Main take home message of experimental design Make sure youve chosen your analysis method and contrasts before you start your experiment!

Why is it so important to correctly design your experiment? Main design goal: To test specific hypotheses We want to manipulate the participants experience and behaviour in some way that is likely to produce a functionally specific neurovascular response.

What can we manipulate? Stimulus type and properties Stimulus timing Participant instructions Adaptation - Repetition suppression

Repeated viewing of the same face elicits lower BOLD activity in face-selective regions Repetition suppression / adaptation designs: BOLD decreases for repetition used to infer functional specialization for this task/stimulus Henson, Dolan, Shallice (2000) Science Henson et al (2002) Cereb Cortex Types of experimental design 1. Categorical - comparing the activity between

stimulus types 2. Factorial - combining two or more factors within a task and looking at the effect of one factor on the response to other factor 3. Parametric - exploring systematic changes in brain responses according to some performance attributes of the task Categorical Design

Categorical design: comparing the activity between stimulus types Example: Stimulus: visual presentation of 12 common nouns. Tasks: decide for each noun whether it refers to an animate or inanimate object. goat bucket

Factorial design combining two or more factors within a task and looking at the effect of one factor on the response to other factor Simple main effects e.g. A-B = Simple main effect of motion (vs. no motion) in the context of low load Main effects LOW e.g. (A + B) (C + D) = the main effect of low LOAD

load (vs. high load) irrelevant of motion HIGH Interaction terms e.g. (A - B) (C D) = the interaction effect of motion (vs. no motion) greater under low (vs. high) load MOTION NO MOTION

A B C D Factorial design in SPM

Main effect of low load: (A + B) (C + D) A B C [1

1 A B C -1

0 D -1 -1] Simple main effect of motion in the context

of low load: (A B) [1 D 0] Interaction term of motion greater under low

load: (A B) (C D) A [1 B -1

C D -1 1] Factorial design in SPM Parametric design

= exploring systematic changes in brain responses according to some performance attributes of the task Parametric designs use continuous rather than categorical design. For example, we could correlate RTs with brain activity.

Overview Experimental Design Types of Experimental Design Timing parameters Blocked, Event-Related & Mixed Design Experimental design based on the BOLD signal A brief burst of neural activity corresponding to

presentation of a short discrete stimulus or event will produce a more gradual BOLD response lasting about 15sec. Due to noisiness of the BOLD signal multiple repetitions of each condition are required in order to achieve sufficient reliability and statistical power.

Design & Neuronal Model Design (Randomized vs. Block) Neuronal Model (Events vs. Epochs) Blocked design = trial of one type (e.g., face image) = trial of another type

(e.g., place image) Multiple repetitions from a given experimental condition are strung together in a condition block which alternates between one or more condition blocks or control blocks Advantages and considerations in Block design The BOLD signal from multiple repetitions is additive Blocked designs remain the most statistically powerful designs for fMRI

experiments (Bandetti & Cox, 2000) Can look at resting baseline e.g Johnstone & colleagues Each block should be about 16-40sec Disadvantages Although block designs are more statistically efficient event related designs often necessary in experimental conditions Habituation effects

In affective sciences their may be cumulative effects of emotional or social stimuli on participants moods Event related design time In an event related design, presentations of trials from different experimental conditions are interspersed in a randomised order, rather then being blocked together by

condition In order to control for possible overlapping BOLD signal responses to stimuli and to reduce the time needed for an experiment you can introduce jittering (i.e. use variable length ITIs) Advantages and considerations in Event-related

design Avoids the problems of habituation and expectation Allows subsequent analysis on a trial by trial basis, using behavioural measures such as judgment time, subjective reports or physiological responses to correlate with BOLD Using jittered ITIs and randomised event order can increase statistical power Disadvantages

More complex design and analysis (esp. timing and baseline issues). Generally have reduced statistical power May be unsuitable when conditions have large switching cost Mixed designs More recently, researchers have recognised the need to take into account two distinct types of neural processes during fMRI tasks 1 sustained activity throughout task (sustained activity)

e.g. taking exams 2 brain activity evoked by each trial of a task (transient activity) Mixed designs can dissociate these transient and sustained events (but this is actually quite hard!) Study design and efficiency

The Basics General linear model: Y = X*+E Where Y is the Matrix of BOLD signals (what you collect), X is the Design Matrix (what you put into SPM), represents the Matrix Parameters (need to be estimated), E represents the error matrix (residual error for

each voxel). Terminology Trials replication of condition. Either epochs: sustained neural activity or events: bursts of neural activity

ITI time between start of one and start of the next trial SOA (stimulus onset asynchrony) time between onset of components. BOLD response The BOLD response to a brief burst of activity typically exhibits a

peak at around 4-6 s and an undershoot at around 10-30 s. To get predicted response Convolve the haemodynamic response with the stimulus. Convolution is a mathematical operation on two functions that produces a third function which typically represents a modified version of one of the original functions.

On timing Fixed SOA of 16 s not particularly efficient. Try much shorter SOA of 4 s IR to events now overlaps considerably. Variability in response is low which means most of the signal will be lost after high pass filtering, so this is not an efficient design, either.

What if we vary SOA randomly? SOA is still 4s, but with a 50% probability of event occurring every 4 s. More efficient because there is larger variability in signal, and we know how the signal varies (even though it is generated randomly, we know this from observing the resulting sequence). Blocked design Runs of events followed by rest periods (periods of null events) blocked design,

very efficient Fourier transform decomposes signal into its constituent frequencies represents signal in frequency space allows us to gain insight into how much of the signal lies within each frequency band Why is it useful?

Take the Fourier transform of each function in the top row, and plot amplitude (magnitude) against Frequency. The neural activity represents the original data, IR acts as a filter (low pass in this case). What is the most efficient design? From what we have seen so far, the most efficient design means varying the neural activity in a sinusoidal fashion with a frequency that matches the peak of the amplitude

spectrum of the IR filter. Sinusoidal modulation places all the stimulus energy at the peak frequency as represented by the single line in the bottom RH corner. High pass filtering We know that there is some noise associated with the scanner. This basically consists of low frequency 1/f noise and

background white noise. We need to filter such that noise is minimised while we keep as much of the signal as possible. For example Consequences of high pass filtering for long blocks. Much of the signal is lost because the fundamental frequency (1/160s ~ 0.006 Hz) is lower than the high pass cutoff. This is why block length should not be too long.

Revisiting our stochastic design Here, the signal is spread across a range of frequencies. Some of the signal is lost due to filtering, but a lot of it is passed which makes this a reasonable design. General linear model revisited Recall:

Y = X*+E Efficiency is basically the ability to estimate given data X and contrast c e (c, X) = inverse (2 cT Inverse(XTX) c) Can only alter c and X Timing differential vs. main effect Differential effect = A-B Optimal SOA (randomised design) = minimal SOA (<2s)

Main effect = A+B Optimal SOA = 16-20s because we are comparing to baseline. Sampling/jitter Jitter is used to randomise SOA Null events can be introduced using jitter Efficient for differential and main effects at short SOA

For SPM Conclusions 1. Do not contrast conditions that are far apart in time (because of low-frequency noise in the data). 2.

Randomize the order, or randomize the SOA, of conditions that are close in time. Also: Blocked designs generally most efficient (with short SOAs, given optimal block length is not exceeded) Think about both your study design and contrasts before you start! References Harmon-Jones, E. y Beer, J. S. (Eds.) (2009). Methods in social neuroscience. Nueva York: The Guilford Press. Johnstone T et al., 2005. Neuroimage 25(4):1112-1123 Previous MfD slides Thanks to our expert Tom Fitzgerald

Recently Viewed Presentations

  • Copyright Notice

    Copyright Notice

    Spielmans GI, Berman MI, Linardatos E, Rosenlicht NZ, Perry A, Tsai AC (2013). Adjunctive atypical antipsychotic treatment for major depressive disorder: a meta-analysis of depression, quality of life, and safety outcomes.
  • Information Technology Project Management, Eighth Edition

    Information Technology Project Management, Eighth Edition

    Learning Objectives, Part 2. Review a case study of an organization applying the project management process groups to manage an IT project, describe outputs of each process group, and understand the contribution that effective initiating, planning, executing, monitoring and controlling,...
  • Focus Districts: Identification, Requirements and Interventions Monday, February

    Focus Districts: Identification, Requirements and Interventions Monday, February

    Next, the 2013-14 & 2014-15 combined Mean Student Growth Percentile (MGP) is checked to see whether any subgroup's MGP is above the state average. The SWD subgroup meets the Progress Filter. The EM Level SWD subgroup PI is removed from...
  • Policy and Implementation Plan for Public Access to

    Policy and Implementation Plan for Public Access to

    Is there a grace period for implementing the new requirements? ANSWER: No. However, ORO will give facilities a reasonable period of time to revise their relevant policies and procedures. Facilities are reminded that requirements for results access took effect on...
  • Apple Presentation Template Supplement

    Apple Presentation Template Supplement

    It provides object-oriented access to database environments WebObjects ? Application Deployment Environment It's an application server environment that sits between a Web server (which interacts with a browser) and data sources. High Level View Advantages ...
  • Diapositivo 1 - Amazon S3

    Diapositivo 1 - Amazon S3

    What's the plural of wolf? WOLFES. WOLVES. WOLFESES. wolves. one wolf. two. GOOD JOB!
  • 15-1 The Puzzle of Life&#x27;s Diversity

    15-1 The Puzzle of Life's Diversity

    SECTION E: VARIATION AND SELECTION OF THE EYGPTIAN ORGAMI BIRD. INTRODUCTION: The Egyptian Origami Bird (Avis papyrus) lives in arid regions of North Africa. Only those birds which can successfully fly the long distances between the sparsely spaced oases will...
  • Created by: Mackenzie Wood - Okaloosa County School District

    Created by: Mackenzie Wood - Okaloosa County School District

    Japanese culture. The Japanese believed in religious and philosophical beliefs. According to their main religion, Shintoism, illness and disease are considered unclean and impure. Therefore, in order to stay disease free, you must show respect for the gods by washing...