VIDEPE

Visualization & Integration of Data for Epilepsy Presurgical Evaluation

A web-based viewer for EEG and neuroimaging data (MRI, PET, SPECT), developed at the Epilepsy and Brain Networks Lab, UNIGE.

No installation needed — just open videpe.github.io/videpe in your browser and start exploring.

Note: VIDEPE is currently under active development. Features and interfaces may change, and some functionality may be incomplete or subject to revision.

Privacy-first: 100% local processing

All data processing happens entirely in your browser. No files are ever uploaded to a server, sent to a third party, or stored outside your own machine. This makes VIDEPE safe to use with sensitive medical data — including identifiable patient recordings — without any institutional data-sharing agreement or de-identification step.

Features

• EEG viewer — high-performance multichannel viewer built on uPlot, optimised for long recordings with many channels. All plots share a single time axis — panning or zooming one channel instantly updates all others.

  • Adjustable range (µV scale), window size, and time step
  • Interactive timeline scrubber for fast navigation across the full recording
  • Keyboard navigation — arrow keys for range/panning, Page Up/Down and Space for window jumps, Home/End to jump to the start/end
  • Configurable number of simultaneously visible channels
  • Min-max downsampling keeps rendering fast at any zoom level
  • Sliding-window buffer — only a portion of the recording is kept in memory at once, so hour-long, high-channel-count recordings stay fast without exhausting browser memory
  • EEG topography — a resizable panel showing a 3D voltage map at the selected time point using NiiVue mesh rendering
    • Re-reference on the fly: none, average, or median reference
    • Diverging blue-white-red colormap with a live colorbar (µV), plus a colour-blind-friendly mode
    • Individual electrode markers on the mesh — voltage-coloured for matched channels, grey for unmapped template positions
    • Uses Standard 10-05 electrode positions by default; load custom positions via .elc or .tsv file
    • 3D rotation is synchronised with the neuroimaging viewer — rotate one and the other follows
  • Intracranial (iEEG) support — toggle between scalp EEG and intracranial (sEEG/ECoG) recordings; iEEG channels are auto-detected from channel naming and shown as a per-electrode voltage matrix, or rendered as a 3D electrode connectome in the neuroimaging viewer once electrode positions are loaded
  • Electrical Source Imaging — When uploading an Inverse Solution file (currently only Fieldtrip *_inversefilters.mat supported), renders per-source power at the selected EEG time point as either a 3D connectome or a volumetric heatmap, with a toggle to switch between the two. Requires the Average montage — VIDEPE switches to it automatically when an inverse solution is loaded, and hides the layer if you switch away.

• Neuroimaging viewer — full multiplanar and 3D rendering powered by NiiVue. Load multiple volumes simultaneously and adjust each one independently.

  • Supports NIfTI (.nii, .nii.gz), MGH/MGZ, GIFTI, PLY, OBJ
  • Multi-layer support for MRI, PET, and SPECT in one view
  • Per-layer opacity, colormap (grayscale, viridis, magma, mako), inversion, and colorbar
  • Drag-to-reorder layers, with a modality subtype label shown on each volume
  • Slice view buttons for axial, coronal, sagittal, multiplanar, and 3D render
  • Switches to auto layout when maximised, restores to grid on collapse

• Built-in demo — hit Load Demo to instantly load a synthetic EEG recording alongside aligned MRI, PET, and SPECT volumes. No upload, no account, no wait.

• Drag & drop file loading — drop files directly onto either viewer panel. VIDEPE detects the format automatically and guides you when multiple files are required.

  • EEG: BrainVision (.vhdr + .eeg) — drop both together or one at a time
  • Volumes: NIfTI, MGH/MGZ, GIFTI, PLY, OBJ
  • Drop multiple imaging files at once to load them as separate layers, or append them to an already-active neuroimaging viewer

• Side-by-side split view — EEG and neuroimaging panels with a draggable divider. Each panel can be independently maximised, restored, or reset, and the two panels can be swapped without reloading any data.

• Cross-platform — runs entirely in the browser, no installation or plugins required. Works on modern desktop and mobile browsers. Dark/light mode follows OS preference, with a manual toggle always available.

• Open source — licensed under AGPL-3.0. The source code is publicly available on GitHub. All derivative works must remain open source. Bug reports, feature requests, and pull requests are welcome.

Supported formats

Modality	Formats
EEG	BrainVision (.vhdr + .eeg)
Electrode positions	.elc, .tsv
Volumes	NIfTI (.nii, .nii.gz), MGH/MGZ, GIFTI, PLY, OBJ

Tech stack

Area	Library
Framework	React 19 + Vite 8
Routing	React Router v7 (HashRouter for GitHub Pages)
Styling	Tailwind CSS v4
EEG plots	uPlot + uplot-react
Neuroimaging	NiiVue
Testing	Vitest + React Testing Library

Prerequisites

VIDEPE uses npm to manage dependencies. npm is bundled with Node.js — installing Node.js is all you need.

Getting started

npm install
npm run dev

The app is served at http://localhost:5173/videpe/.

Running tests

npm test          # watch mode
npm run test:run  # single run (used in CI)

Deployment

The app deploys automatically to GitHub Pages on every push to main. Tests run on every pull request.

Links

• GitHub repository
• UNIGE Open Science page