Switch branches/tags
Nothing to show
Find file History
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
..
Failed to load latest commit information.
Freeze
model
modelWB
notebooks
output
outputWB
MNI3mm.nii.gz
MNI_GMmask3mm.nii.gz
README.md
runClassifier.py

README.md

ML

Classifier Design: HoJin, Swati, Ann, Eshin https://github.com/preprocessed-connectomes-project/abide

Available fMRI: 3D volume per subject, redo Structure: Freesurfer (Desikan Killany) Demographic data: Phenotypic_V1_0b_preprocessed.csv: (ABIDE_LEGEND_v1.02.pdf) 73 potential features

Steps Look at run.py Download data - works, downloaded reho data only from PITT for starters python download_abide_preproc.py -d reho -p cpac -s filtnoglobal -o reho -t PITT

Mini Project 1: Read csv file of phenotypic information - removed QA measures from ABIDE, will use from our group Reduce data based on csv (remove not research reliable data) SEX (474M-65F/474M-99F), AGE, DX, FIQ values are available for most subjects ~ 900, otherwise dataset reduces to 200 Do PCA or some data reduction - not needed

Simple explanation of ReHo: http://nro.sagepub.com/content/early/2015/10/28/1073858415595004.full

Mini Project 1 apply to cortical measures Mini Project 1 apply to fMRI voxelwise data

Mini Project 2: Combine reduced features (phenotypic, structural, fMRI) Run Miniproject 1 of the combined features

Mini Project 3: Classifier: What kind?

Extra Mini Project 4: Compare with Deep Learning algorithm on reduced datasets