Analyzing EEG recordings rigorously with MNE-Python and benchmarking decoders with MOABB typically starts with hours of Python environment setup, dependency management, and boilerplate scripting. With the launch of PiEEG Studio, that entire scientific workflow now runs directly in your browser—and writes clean, reproducible Python code for you as you work.
Accessible standalone or directly from PiEEG Cloud (via the Session Lobby, Saved Sessions, or playback toolbar), PiEEG Studio accepts CSV, Excel, EDF, BDF, FIF, and live browser captures with zero local installation.
Full 11-View MNE-Python Pipeline
Drop in a recording—or generate a synthetic paradigm—to move interactively through a complete electrophysiology workflow:
- Signal Browser & Channel Quality: Inspect multi-channel traces, bad-channel metrics, and montage layouts.
- Preprocessing & Artifact Removal: Apply zero-phase bandpass/notch filters, common average or custom re-referencing, and Independent Component Analysis (ICA).
- Spectral & Topographic Mapping: Compute Welch/multitaper power spectral densities alongside 2D scalp topomaps across canonical frequency bands.
- Epochs, ERPs & Time-Frequency: Slice event-locked epochs, inspect evoked responses, and visualize time-frequency spectrograms.
- Functional Connectivity: Evaluate cross-channel coherence and phase-coupling matrices.
Built-In MOABB BCI Benchmark Lab
The BCI & MOABB workspace brings standardized decoding benchmarks—including Motor Imagery, P300/ERP, and SSVEP—straight into the browser across 7 classic and Riemannian pipelines:
CSP + LDA(Common Spatial Patterns + Shrinkage LDA)CSP + SVM(Common Spatial Patterns + RBF SVM)Riemannian TS + LR(SPD Covariance + Tangent Space + Logistic Regression)Riemannian MDM(Minimum Distance to Riemannian Mean)LogVar + LDA(Channel Log-Variance + Shrinkage LDA)BandPower + RF(Welch Band Powers + Random Forest)ERP-Xdawn + LDA(Xdawn Spatial-Temporal Covariance + LDA)
Compare accuracy, balanced accuracy, F1, and ROC-AUC against chance level, inspect confusion matrices and CSP spatial pattern topomaps, and validate within-session or cross-session generalization.
Every Click Becomes Reproducible Python Code
PiEEG Studio is not a closed GUI box. Every filter, montage change, epoch window, and MOABB benchmark you configure automatically generates a standalone Python script or Jupyter notebook that you can download, version-control, customize, and batch-run on new datasets.
In the Cloud Preview, each browser session is strictly isolated via per-session sandboxing (X-Studio-Session) with automatic 15-minute inactivity cleanup; if you run the local Community Edition (127.0.0.1:1618), PiEEG Cloud automatically routes Studio actions locally for unlimited file sizes and 150+ online MOABB datasets.
