Back to Projects

Precision Neural Engineering Lab

ePOCHE

Visual EEG-to-model pipeline

Lead Developer · 2023–Present · Open Source

Python PyQt6 SQLite scikit-learn TensorFlow MNE-Python NumPy SciPy Matplotlib

ePOCHE is a GUI-driven time-series classification platform that takes you from raw EEG to trained models—with feature engineering, model management, and automated evaluation—all through a visual interface.

Built for researchers who need to iterate on classification approaches without rebuilding infrastructure every time.

View on GitLab →

ePOCHE is designed as a domain-specific IDE for biosignal classification: visual pipelines, systematic comparison, and reproducibility by default.

Hover to playTap to play

ePOCHE — Visual EEG-to-model classification pipeline

Platform Architecture

ePOCHE Classification Platform
Experiment Manager
Session isolation · Full state tracking (SQLite) · Reproducible from any checkpoint
Signal Processing Pipeline
Multi-format ingestion (EDF, BDF, CSV) · Preprocessing · Quality metrics · Artifact detection
Feature Engineering
Time-domain · Frequency-domain (FFT, PSD) · Time-frequency (wavelets) · Configurable extraction
Classifier Library (11 Types)
Logistic Regression Random Forest SVM Neural Networks XGBoost Decision Trees KNN Naive Bayes AdaBoost Gradient Boosting MLP
Results & Reporting
Cross-validation metrics · Confusion matrices · Feature importance · Full reproducibility

ML Pipeline Flow

Data Loader
Feature Extract
Model Training
Cross-Validation
Export Results
Production Use: Active NIH R01 study (Parkinson's research) · 11 classifier types · Database-driven reproducibility

What Problem It Solves

Most EEG classification workflows assume:

  • You'll write throwaway scripts for each experiment
  • You'll reinvent preprocessing, feature extraction, and evaluation every time
  • Reproducibility is your problem—buried in ad-hoc code
  • You have deep ML engineering knowledge (or will acquire it)

This creates fragile pipelines, inconsistent results, and weeks lost to rebuilding infrastructure that should exist once and work everywhere.

ePOCHE replaces that chaos with a visual pipeline you can configure, run, and reproduce.

What Makes ePOCHE Different

Visual Configuration

  • GUI-driven pipeline — configure the entire workflow visually
  • No code required — researchers can iterate without engineering support
  • Reproducible by design — every run captures its exact configuration

Signal Processing

  • Multi-format ingestion — EDF, BDF, CSV supported out of the box
  • Preprocessing pipeline — filtering, resampling, epoch extraction
  • Signal quality metrics — automated artifact detection before you waste compute

Feature Engineering

  • Time-domain features — statistical measures across epochs
  • Frequency-domain features — FFT, power spectral density
  • Time-frequency features — wavelet transforms for non-stationary signals
  • Configurable extraction — select exactly what matters for your task

Model Management

  • Unified API — scikit-learn and TensorFlow models through the same interface
  • Automated cross-validation — configurable strategies without boilerplate
  • Hyperparameter optimization — grid and random search built in
  • Comparative visualization — evaluate model performance side by side

Export & Reproducibility

  • Trained model export — save and reload models
  • Feature importance reports — understand what drives predictions
  • Reproducibility logs — every parameter captured for publication

Designed for How Researchers Actually Work

ePOCHE is not just a script collection—it is a thinking environment for biosignal classification:

  • You explore data before committing to a model
  • You iterate on features without rewriting pipelines
  • You compare approaches systematically, not ad-hoc
  • You export results that reviewers can reproduce
Domain experts should configure behavior through interfaces, not source code. The same design philosophy that drives the MUSE ecosystem.

Why This Exists

ePOCHE grew out of watching researchers rebuild the same infrastructure for every EEG study—loading data, extracting features, training models, generating figures. The exact pipeline was always buried in scripts that worked once and broke forever.

The abstraction layer means the same tool works for EMG, ECG, GSR—any biosignal classification. Swap data sources and model types without touching underlying code.

Platform lesson: Researchers don't want to rebuild the same ML pipeline infrastructure for every study—they want to iterate on classification approaches. The same way developers don't want to rebuild CI/CD for every project—they want consistent, reproducible workflows. ePOCHE is a control plane for biosignal classification.

Impact

  • Reduced pipeline setup from days to hours
  • Enabled systematic comparison of classification approaches
  • Made results reproducible by default
  • Accessible to researchers without ML engineering background

Who It's For

  • Cognitive neuroscience researchers
  • BCI developers prototyping classification
  • Graduate students learning ML for biosignals
  • Anyone who needs reproducible EEG/EMG/ECG analysis