Ted Kelly

Back to Psych 204 Projects 2009

Multi Voxel Pattern Analysis in the Ventral Temporal Cortex

Recent Studies have shown that there are clearly defined regions within the ventral temporal cortex that preferentially respond to faces (fusiform face area), places (Parahippocampal Place Area) and objects. In particular, the FFA has been shown to undergo development from childhood throughout adolescence. This study uses multi voxel pattern analysis to confirm both the location and the growth of the FFA in the VTC.

Background

Multi Voxel Pattern Analysis vs. ROI analysis

Most of the current research on the VTC and the FFA has focused on region of interest (ROI) analyses. This analysis focuses on the highest activations of the voxels selective for faces. Usually, activations are thresholded so that only voxels responding over a certain level are looked at. Supporters of this type of analysis say that the highest activations are the most interesting and provide the clearest evidence for the age related growth of the FFA (in volume and in more well-defined areas selective for faces vs. non-face stimuli). However, multi voxel pattern analysis (MVPA) supporters insist that there is much more information in all the activations in the VTC, not just the highest ones. They say that the activations are distributed throughout the entire VTC and that the aforementioned growth of the FFA is actually just growth of the entire VTC. This study aims to prove the MVPA supporters wrong by using MVPA to support the initial analyses provided by ROIs.

Methods

Subjects

Multi-Voxel Pattern Analysis was done on 14 adolescents (ages 12-16yrs) and 11 adults (ages 18-40yrs)

MR Acquisition

Brain imaging was performed on a 3 tesla whole-body General Electric Signa MRI scanner. During fMRI subjects viewed gray-scale images of the following types: faces of male children and adolescents (ages 6–16 years), faces of male adults (ages 18–40 years), abstract sculptures, cars, indoor scenes, outdoor scenes, and scrambled images (created by randomly scrambling pictures into 225, 8 × 8 pixel squares). Stimuli were presented in 12 s blocks followed by 12 s of a blank screen with a fixation at a rate of 1 Hz. Subjects participated in two 396-s runs with different images.

MR Analysis

Below is an example of an initial MVPA analysis on the VTC. The MVP reflects differences in category selectivity in each voxel, rather than amplitude differences across voxels. For each stimulus type, MVPs were generated across the anatomical ROI of VTC, separately for data from run 1 and run 2. Corresponding graphs reflect the within-category reproducibility and across-category distinctness of MVPs to determine if there are between age group differences that extend beyond functionally defined ROIs.

Figure 1

From these initial MVP results we decided to create new ROIs by dividing the VTC in two. We created lateral and medial VTC ROIs for each subject using MrVista. We used the middle fusiform sulcus as the anatomical landmark to divide our ROIs.

Pre-processing

No thresholding or spatial smoothing was applied to the MVPA data. Using Z-scores to calculate correlations minimizes between-voxel effects

Results

Lateral VTC scans

The Lateral VTC scans show good reproducibility of patterns within category and show activation in the lateral part of the lateral VTC in response to faces where previous ROI analysis has shown the FFA is located. These scans are of a 14 yr old boy.

Figure 1

Lateral VTC graphs

These graphs confirm the good reproducibility within category for all subjects both adolescents and adults. Reproducibility is better for faces and scenes than objects. Between category correlations suggest that there are no age related differences in age of face effects. Adults have better ability to distinguish between faces and cars, faces and absobj and faces and scenes. This also means that the areas of activation are more distinct in adults vs adolescents.

Figure 1

Retinotopic models in individual subjects transformed into MNI space

Some text. Some analysis. Some figures.

Retinotopic models in group-averaged data on the MNI template brain

Some text. Some analysis. Some figures. Maybe some equations.

Equations

If you want to use equations, you can use the same formats that are use on wikipedia.
See wikimedia help on formulas for help.
This example of equation use is copied and pasted from wikipedia's article on the DFT.

The sequence of N complex numbers x₀, ..., x_N−1 is transformed into the sequence of N complex numbers X₀, ..., X_N−1 by the DFT according to the formula:

$X_{k}=\sum _{n=0}^{N-1}x_{n}e^{-{\frac {2\pi i}{N}}kn}\quad \quad k=0,\dots ,N-1$

where i is the imaginary unit and $e^{\frac {2\pi i}{N}}$ is a primitive N'th root of unity. (This expression can also be written in terms of a DFT matrix; when scaled appropriately it becomes a unitary matrix and the X_k can thus be viewed as coefficients of x in an orthonormal basis.)

The transform is sometimes denoted by the symbol ${\mathcal {F}}$, as in $\mathbf {X} ={\mathcal {F}}\left\{\mathbf {x} \right\}$ or ${\mathcal {F}}\left(\mathbf {x} \right)$ or ${\mathcal {F}}\mathbf {x}$.

The inverse discrete Fourier transform (IDFT) is given by

$x_{n}={\frac {1}{N}}\sum _{k=0}^{N-1}X_{k}e^{{\frac {2\pi i}{N}}kn}\quad \quad n=0,\dots ,N-1.$

Retinotopic models in group-averaged data projected back into native space

Some text. Some analysis. Some figures.

Conclusions

Here is where you say what your results mean.

References - Resources and related work

References

Software

Appendix I - Code and Data

Code

File:CodeFile.zip

Data

zip file with my data

Appendix II - Work partition (if a group project)

Brian and Bob gave the lectures. Jon mucked around on the wiki.