Gradient Biotech

Data upload

Upload data from the study home page (/areas/neurology/studies/{id}). Each upload creates one dataset and redirects to the dataset workspace.

Upload flow

  1. Select file — drag-and-drop or browse
  2. Import — writes file to storage and creates the dataset record
  3. Redirect — opens the dataset workspace for pipeline actions

Supported formats

ModalityFormatFilesNotes
Connectivity matrixCSVSingle .csvSquare n×n numeric matrix
Connectivity matrixNPZ / NPY.npz or .npyArray key matrix or sole array
Connectivity matrixMATLAB.matVariable matrix, conn, or W
EEG / MEGEDF.edfEuropean Data Format recordings
EEGBrainVision.vhdr + .eegUpload header; companion files detected
EEGEEGLab.setMATLAB EEGLab structure
fMRINIfTI.nii or .nii.gzPreprocessed BOLD timeseries or connectivity input
fMRIBIDSZIP archiveBIDS dataset bundle for validation and FC pipelines

Validation rules

  • Connectivity matrices must be square and symmetric (within tolerance)
  • Non-finite values (NaN, Inf) are rejected or flagged
  • EEG uploads require detectable channel count and sampling rate
  • BIDS ZIP must contain a valid dataset_description.json and subject/session structure

After upload

The dataset appears on the study home page. Click it to open the dataset workspace and run the appropriate pipeline for its modality.

For connectivity matrices, Compute graph metrics is available immediately after upload.

Group labels

Assign group labels from the Cohort analysis page after upload, or include group metadata when importing subjects. At least two groups with multiple datasets each are required for NBS.

Sample data

See Sample data for test fixtures.

For full format specifications, see File formats.