danielzamorap/el4203-proyecto-progra-avanzada-eeg

How danielzamorap/el4203-proyecto-progra-avanzada-eeg works

This repository is a research-oriented EEG seizure-window classifier. Its workflow is to prepare patient-specific windows from CHB-MIT EDF records using seizure annotations, save channel-power features and labels as NPZ files, then load five patients for training and hold out one patient for evaluation. The two models produce seizure scores that metrics summarize and the MLP uses to prioritize review windows; the script also saves an ROC plot. The graph reflects the shown entry points and module calls; it does not assume unsampled notebook or raw-data details.

Main entry points: main.py Train-test runner · metricas.py Model metrics · preparar_datos.py Patient preparation · datos.py Patient datasets

Components 14

Data preparation

Seizure annotations src/anotaciones.py
Patient preparation src/preparar_datos.py

Dataset handling

EEG window features src/datos.py
Patient datasets src/datos.py
Processed NPZ files data/processed

Modeling and evaluation

Train-test runner src/main.py
Threshold model src/modelos.py
Neural network src/modelos.py
Score sorting src/ordenamiento.py
Model metrics src/metricas.py

Review outputs

Prioritized windows src/main.py
ROC plot resultados

Other components

Analyst
CHB-MIT EDF

Connections 19

Analyst → Train-test runner: runs README.md
Analyst → Patient preparation: prepares data src/preparar_datos.py
CHB-MIT EDF → Patient preparation: supplies records README.md
Patient preparation → Seizure annotations: reads annotations src/preparar_datos.py
Patient preparation → EEG window features: adds records src/preparar_datos.py
EEG window features → Processed NPZ files: saves features src/datos.py
Patient datasets → Processed NPZ files: loads patients src/datos.py
Train-test runner → Patient datasets: loads patients src/main.py
Train-test runner → Threshold model: trains and scores src/main.py
Train-test runner → Neural network: trains and scores src/main.py
Train-test runner → Model metrics: evaluates models src/main.py
Model metrics → Patient datasets: reads labels and features src/metricas.py
Model metrics → Threshold model: requests predictions src/metricas.py
Model metrics → Neural network: requests scores src/metricas.py
Neural network → Score sorting: orders scores src/modelos.py
Train-test runner → Patient datasets: gets window origins src/main.py
Train-test runner → Neural network: prioritizes windows src/main.py
Train-test runner → Prioritized windows: prints top windows src/main.py
Train-test runner → ROC plot: saves ROC plot src/main.py

Drawn by GitDiagram on November 13, 2026 from the repository's file tree, README and source files. Read as Markdown ·