How ding1119/neurotree works
NeuroTree is a research pipeline for decoding fMRI brain networks, classifying disorders or estimating brain age, and interpreting predictions through hierarchical brain-tree visualizations. The main workflow loads and transforms dataset time series, trains a Neural ODE graph model, evaluates predictions, extracts trunk paths, and optionally plots them. The README also describes convergence analysis and age-related network interpretation, but the sampled source does not establish their runtime wiring. Edges reflect shown calls or documented usage; dataset placement is documented, while unsampled subsystem details are not inferred.
Main entry points: training_eval_utils.py Training and evaluation · main.py Training entry · load_dataset.py Dataset loading · tree_plot.py Brain-tree plots
Components 11
Data preparation
- Dataset loading
data_handler/load_dataset.py - Connectivity preprocessing
data_handler/load_dataset.py - Input tree construction
utils.py
Modeling and evaluation
- Training entry
main.py - Training and evaluation
training_eval_utils.py - Neural ODE model
models/ode_model.py
Interpretation
- Trunk extraction
tree_trunk_utils.py - Brain-tree plots
visualization/tree_plot.py - Network plotting
visualization/plot_utils.py
Other components
- Researcher
- Processed fMRI data
Connections 10
- Researcher → Training entry: runs
README.md - Processed fMRI data → Dataset loading: loads
README.md - Training entry → Dataset loading: loads data
main.py - Dataset loading → Input tree construction: constructs trees
data_handler/load_dataset.py - Training entry → Training and evaluation: trains and evaluates
main.py - Training and evaluation → Neural ODE model: calls model
training_eval_utils.py - Training and evaluation → Trunk extraction: extracts trunks
training_eval_utils.py - Training and evaluation → Brain-tree plots: plots brain trees
training_eval_utils.py - Brain-tree plots → Network plotting: renders comparisons
visualization/tree_plot.py - Training and evaluation → Network plotting: plots distributions
training_eval_utils.py