NEU 502B Final project - Multi langauge perception with fMRI data

Group members: Will Long, Iman Wahle, Dexter Tsin, Yousuf El-jayyousi, Sae Yokoyama

Analysis code for NEU 502B final project.

Checklist

• code/analysis/preprocess_notebook.ipynb

  • Head motion error
  • acompCor (n_components = 5)
  • Framewise displacement (threshold = 0.25)
  • Confounds array at: /jukebox/PNI-classes/students/NEU502/2023-NEU502B/wlong/multilang/preprocessed_data/confounds.npy

• code/analsis/stim_notebook.ipynb

  • Basic block stimulus with glover hrf
  • block stimiulus with specific language
  • Stim array at: /jukebox/PNI-classes/students/NEU502/2023-NEU502B/wlong/multilang/preprocessed_data/stim_arr.npy

• code/analysis/GLM_notebook.ipynb

  • Basic OLS regression with confounds
  • Basic OLS regression with time-lagged stimulus
  • Regression analysis with FirstLevelModel from nilearn

TODO

• Intro text: Sae

• Results text: Will

• figure 1: Sae

  • figure of task

• figure 2: Yousuf

  • anatomy for both subjects
  • auditory mask
  • mean bold signal in auditory cortex
  • preprocessing (head motion, etc)

• figure 3: Iman

  • GLM coef on brain for familiar language vs not
  • GLM coef on brain for first language vs second

• figure 4: Dexter

  • Word embeddings (optional)