Window by window

Quick start

1. In the launcher, click Set folders and choose your Import folder (raw data) and Export folder (results). You can use one folder for both.
2. New here? Open ? Help (top right), pick a window and click Try it with demo data: it opens that window with synthetic data whose answers are known, so you can practise and check your numbers before using your own recordings. The Course and the Virtual lab are coming soon.
3. Under Analyses, find the tile for your data (grouped as Blood flow, Electrophysiology, EEG, Imaging and Across techniques) and click its numbered steps in order.
4. In every window, work down the numbered step cards on the left: Load → Settings → Run → Save. The teal button is the recommended next action; greyed-out buttons are not possible yet.
5. Read the status bar at the bottom of each window: it says what happened and what to do next.
6. Stuck? Click ? Help in the window header for its quick start and troubleshooting.
7. Many files with the same settings? Use Batch processing in the launcher. To keep an analysis, click Save session… (and Report (PDF)…) in the last step of any window.

Quick start

1. 1 Load recording: click Load file... and choose the recording. Check LDF (the flow channel) and Stimulus (a trigger channel, the comments / event markers of the file, or None); both are guessed from the channel names. With several LabChart blocks, choose the Block. The file label shows the format, rate, duration and number of channels.
2. 2 Choose time range: type Start and End (s), click Pick on plot and click twice (start, end) on either plot, or click Full range.
3. 3 Crop: click Crop to range. The cropped signals replace the full recording in the plots.
4. 4 Save: click Save cropped data... and choose a file name. Open this file next in LDF Process.
5. Session / report (optional): in step 4, Save session… stores the recording (with checksum), the channels, the range and the crop; Open session… redoes them; Report (PDF)… writes a one-page summary. See Sessions and reports.

Quick start

1. Load cropped LDF: click Load file... and choose the file saved by LDF Extract.
2. Filter / downsample (optional): click Settings..., choose downsampling and filter, click Apply. The Filter response tab shows the filter; Undo returns to the loaded data. See the Filtering topic.
3. Segment trials: set Stim threshold (shown as a dashed line on the stimulus), Pre-onset and Post-onset (s) and Min interval (s), then click Segment trials. All trials and the mean ± SD appear in the Trials tab; the Checks tab lists what was checked in the recording and the settings (click a row for why it matters and what to try).
4. Save trials: click Save trials.... Choosing an existing trials file appends the new trials to it (time axes must match). Open the file(s) next in LDF Average.
5. Session / report (optional): in step 4, Save session… stores the file (with checksum), filter and segmentation settings; Open session… re-runs them; Report (PDF)… writes a one-page summary. See Sessions and reports.

Quick start

1. In LDF Process, click Settings (step 2).
2. Choose Downsample first: it sets the new sampling rate and therefore the highest usable cutoff (Nyquist = half the rate).
3. Pick the filter type and enter the cutoff(s) in Hz; start with order 4.
4. Click Apply and compare the filtered trace with the original in the plots; use Undo to try other settings.

Quick start

1. Load trial files: click Add files... and select one or more files saved by LDF Process (multi-select). You can add more files later; Clear all starts over.
2. Options: tick Relative to baseline to subtract each trial's pre-stimulus mean.
3. Grand average: click Plot grand average to see the mean ± SD across all trials.
4. Session / report (optional): in step 3, Save session… stores every trial file (with checksum) and the option; Open session… reloads them and redraws the average; Report (PDF)… writes a one-page summary. See Sessions and reports.

Quick start

1. 1 Load images: click Load images… and choose a .mat (frames H × W × N), a multi-frame TIFF or a video (or Try demo data). Check Images are: Raw speckle images straight from the camera, Speckle contrast (K) images, or Perfusion / flux images already computed by a commercial system (used as they are). Check the Frame rate and, for the 1/τc model, the Exposure (read from the file when it says).
2. 2 Contrast and flow (raw images): Spatial contrast in a 7 × 7 window is the usual choice; Frames per value averages the contrast of several frames (set to about 2 values per second when you load). Type the camera Dark level (an image with the laser off). Flow index: 1/K² (the speckle flow index), or 1/τc from the exposure model.
3. 3 ROIs: click Add ROI and drag a rectangle over each region (the ROIs of a .mat file are loaded with it). Without ROIs the whole image is one region.
4. 4 Stimulus, trials and Run: Onsets from the stimulus trace in the file, a Regular protocol (first onset, interval, count) or None. Before / After set the trial (the part before 0 s is its baseline) and Resp. from / to the response window. Click Run.
5. Read the results: Show → Response map (%) colours where the flow rose (orange) or fell (blue); Average response has one curve per ROI (mean ± SD) with its mean change in the response window in the legend; Flow over time the whole recording; Checks lists what to look at before using the numbers (click a row to read why it matters and what to try).
6. 5 Export: Export results… writes a .csv (time, flow change % and flow index per ROI) or a .mat (maps, traces, trials, masks, settings); Save trials… writes the trials of the selected ROI in the LDF trial format, which LDF Average and Response features open. Save session…, Report (PDF)… and Methods text… keep the analysis.

Quick start

1. Load recording: choose the Source (the acquisition system or file format; see Inputs), click Load recording… and select the file or folder it asks for (a TDT tank / block folder, an Open Ephys or Neuralynx folder, or a data file). The channel lists are filled from the recording.
2. Choose channels: pick the Stimulus channel (TDT: Whis; Intan: DIGITAL-IN / ANALOG-IN, and the stimulation current in .rhs files; Open Ephys: TTL line / ADC; NWB: stimulus TimeSeries or trials; SpikeGLX: sync or digital bits, analog inputs; Blackrock: analog inputs, digital input bits from the .nev; Neuralynx: TTL bits) and one or more raw channels (All / None help).
3. Process: optionally Plot RAW; then Process LFP… (low-pass, optional 60 Hz notch, downsample) and/or Process MUA… (band-pass, default 300–3000 Hz).
4. Save: type the probe's Electrode spacing (µm) if you know it (0 = not known; it is saved with the LFP and used for the CSD), then Save LFP… / Save MUA…, choose which channels to keep and a file name (default <recording>_LFP.mat / <recording>_MUA.mat). Open these files in LFP analysis / MUA analysis. Export NWB… writes the processed LFP and its stimulus channel as an NWB 2.x file (default <recording>_LFP.nwb).
5. Session / report (optional): in step 4, Save session… stores the recording (with checksum), channels and LFP / MUA settings; Open session… re-processes them; Report (PDF)… writes a one-page summary. See Sessions and reports.

Quick start

1. Load LFP file: click Load LFP file… and choose the LFP file saved by Extract Ephys.
2. Channels: select the channels to analyse (All / None).
3. ERP analysis: click Run ERP…, set pre- and post-stimulus time (s), stimulus threshold and minimum ISI (s), click OK. The number of averaged epochs is reported; the Checks tab lists what was checked (click a row for why it matters and what to try).
4. CSD: choose the Method (Standard, iCSD delta, iCSD step, iCSD spline or kCSD), enter Spacing (µm) (filled in from the file when it gives lfp_spacing_um; otherwise type your probe's contact spacing, the default is 100 µm) and the Channel order from top to bottom (at least 3 channels of the last ERP), then click Compute CSD; the Checks tab then also checks the spacing and where the CSD sink lies. iCSD and kCSD also ask for the conductivity σ (S/m) (0.3 for cortex) and the Diameter (µm) of the active tissue (500 by default). iCSD has an optional Smoothing (µm) (0 = off). kCSD has R (µm) and λ, where 0 = chosen automatically by cross-validation. Only the fields of the chosen method are shown (which method?).
5. Export: click Export ERP / CSD… to save a .mat that Signal Characterization can read.
6. Time–frequency: choose the Channel, Frequencies (Hz) (lowest – highest), Wavelet cycles, Epoch (s) and Baseline (s). The band table (delta 1–4, theta 4–8, alpha 8–13, beta 13–30, gamma 30–80 Hz) holds common conventions: edit the limits for your preparation and tick Plot for the bands to show. Try oscillation demo loads a demo with known theta and gamma oscillations and selects channel 4.
7. Time–frequency plots (step 6): click Spectrum (power spectrum of the whole recording), Spectrogram (power over time with the stimuli marked), ERSP / ITPC (power change in dB and phase locking around each stimulus) and Band power (% change of each ticked band around the stimulus, mean ± SEM). Each opens its tab. ERSP and Band power use the stimulus onsets found with the ERP threshold (0.5 until you run the ERP).
8. Session / report (optional): in step 5, Save session… stores the file (with checksum) and the ERP, CSD and time–frequency settings and results; Open session… re-runs them; Report (PDF)… writes a one-page summary. See Sessions and reports.

Quick start

1. Load MUA file: click Load MUA file... and choose the MUA file saved by Extract Ephys. Channels and whether a stimulus is present are shown.
2. Channel & segments: choose the channel; optionally tick Segment by stimulation onsets (minimum ISI, threshold, pre / post-stimulus times) and pick a segment.
3. Spike sorting: click Configure... (detection method, threshold, polarity, filtering, features, clustering, Auto-merge similar clusters with its Merge threshold (correlation), drift correction) and then Run. With auto-merge on (default), clusters whose mean waveforms have the same shape (correlation ≥ 0.95) and size (amplitude ratio ≥ 0.85) are joined; the status bar says what was merged.
4. Clusters: select the clusters to show (Select all / Clear) and read the quality summary. Select two or more units and click Merge selected when they are the same neuron; select one unit and click Split selected to cut it in two; Undo reverses the last merge, split or auto-merge. The Checks tab lists what was checked on the sorted channel (refractory period, signal-to-noise, amplitude drift; click a row for why it matters and what to try) and follows every merge, split and Undo.
5. Raster & PSTH tab: spikes of each selected unit (up to 4) around every stimulus onset (raster) and the mean firing rate ± SEM (PSTH). Set From (s), To (s) and Bin (ms); onsets use the threshold and minimum ISI of Segment by stimulation onsets (default 0.5, 1 s).
6. Correlograms tab: autocorrelograms (diagonal) and cross-correlograms of up to 4 selected units; set Max lag (ms) and Bin (ms). The shaded band is ± the refractory period: a clean unit has (almost) no spikes there.
7. Export: click Save results... to write spike times, cluster IDs, the sorting parameters, quality measures and the list of merges / splits (info.clusterEdits) to .mat.
8. Session / report (optional): in step 5, Save session… stores the file (with checksum), all settings and the sorted clusters with your edits; Open session… restores them as saved (Undo still works); Report (PDF)… writes a one-page summary. See Sessions and reports.

Quick start

1. Load stack: click Load stack and choose a .mat or multi-frame TIFF (or Try demo data / Try advanced demo (motion, 3 cells)). The first frame is shown with the size, frame count and time source; a roiMask / roiMasks in the file becomes the first ROI(s).
2. Preprocess (optional): tick Motion correction (rigid) if the frames jitter: the shifts are estimated once (max shift shown next to the box, per-frame plot in the Motion correction tab) and applied before everything else. Then optionally B&W 256 levels, Smooth and/or Normalize each frame, applied in that order when you click Run.
3. ROIs and line: click Add ROI and drag a rectangle (repeat for more ROIs), or Detect cells to add one ROI per active cell automatically. ROIs are listed next to the image (double-click a name to rename, Remove ROI to delete). For line methods click Draw line (across the vessel for diameter). Clear ROIs and line starts over.
4. Analysis: choose the Method (for ΔF/F also the baseline frames; for Vessel diameter optionally Robust diameter (ignore blood cells)) and click Run. ROI methods give one trace per ROI, in the ROI's colour. The Checks tab (next to Result and Motion correction) lists what was checked: movement against the size of the cells, bleaching and clipped pixels; click a row for why it matters and what to try. The status bar says when there are warnings.
5. Export: click Export results to save a .csv (time + one column per measure and ROI) or a .mat (all series, ROI masks and names, line, shifts and settings).
6. Session / report (optional): in step 5, Save session… stores the stack (with checksum), ROIs, line and settings; Open session… re-runs motion correction and the analysis; Report (PDF)… writes a one-page summary. See Sessions and reports.

Quick start

1. 1 Load images: click Load images… and choose one or more files (or Try demo data). TIFF: every page is a channel; PNG / JPG: the colours are channels; .mat: variable images (H × W × channels × images). Several files = several sections, time points or wells, with the same channels. Check the Pixel size (µm): it is read from ImageJ TIFFs; otherwise type the value of your microscope.
2. 2 Align (optional): tick Align channels if the colours of one cell do not sit on top of each other. With several images choose Align images: Shift (automatic) for the same field imaged again (time points), Landmarks (click points) for serial sections: click Pick landmarks…, click the same 3 or more features in image 1 and in the image shown, then Finish (Clear landmarks starts again). Click Align and switch images to check the overlay.
3. 3 Count cells: choose the channel in which every cell is visible (usually nuclei). The defaults suit nuclei of about 8–12 µm; click Count cells. Circles mark counted cells, grey × marks objects that were not counted. Switch Show to What was thresholded to see what counted as cell.
4. 4 Regions and markers: every other channel is a marker; a cell is positive when at least Positive if (% of cell) of it is bright in that channel (yellow circles = positive for every marker). Click Add region, click the corners on the image, then Finish, to get counts and densities per region. Import regions… adds regions drawn in ImageJ / Fiji (a .roi, or the ROI Manager's RoiSet.zip) or QuPath (annotations exported as GeoJSON), in the pixels of image 1; lines and points are skipped.
5. 5 Export: Export results… writes a .csv with one row per cell plus <name>_counts.csv (per image and region), or a .mat with everything. Save session…, Report (PDF)… and Methods text… keep the analysis.
6. Read the Checks tab before using the numbers: it says where the pixel size came from, how well the images and channels were aligned, which threshold was used and what was left out. Click a row to read why it matters and what to try.

Quick start

1. 1 Load EEG: Load EEG files… (one file per participant: EEGLAB .set, FieldTrip .mat, BrainVision .vhdr, EDF / BDF, XDF, EEG-BIDS or a plain .mat), Try demo data (8 cleaned participants) or Try raw demo (continuous, not cleaned) (3 raw recordings).
2. Read the Overview tab: channels, trials per condition, reference, electrode positions and what was already done to the data, then every step done here.
3. Electrode layout… (step 1): check the drawing and the Status column, fix the Source, a Positions file… or a name in As if needed, and click Use this layout. Positions are needed only for scalp maps.
4. 2 Clean recordings (raw data): Bad channels per participant (or Suggest), High-pass, Low-pass and Notch filters, the Reference, then Apply.
5. 3 Trials: Events and names (e.g. S 1 = Standard, S 2 = Target), Trial from / to (ms), Reject trials by Peak-to-peak and/or Absolute amplitude, then Cut into trials (or Apply rejection for files already cut into trials). The Checks tab lists what was checked in the trials (click a row for why it matters and what to try; Show ERPs and Measure update it).
6. 4 ERPs: baseline (−200 to 0 ms by default), Channels to plot (e.g. Pz), Show ERPs; choose the participant or grand average and the view above the plot.
7. 5 Measure: mean amplitude (recommended) or peak amplitude, the time window and channels, then Measure (Measures tab).
8. Scalp maps (step 5, next to Measure): the mean voltage of every electrode in the same window, drawn on the head (or on the skull for rodents), one map per condition and one of A minus B; confirm the electrode layout first. Show → Conditions brings the ERPs back.
9. 6 Statistics: Parametric or Nonparametric, Compare conditions (Statistics tab; the Checks tab adds the checks of the test).
10. 7 Time–frequency (optional): Frequencies (e.g. 4 to 40 Hz), Wavelet cycles, Baseline, Channels (e.g. Oz) and Band, then Time–frequency; Show chooses Time–frequency (ERSP), Phase locking (ITPC) or Band power. Grey = no value (the wavelet would reach beyond the trial).
11. 8 Save: Export results… (.csv or .mat), Save session… (it replays every cleaning step from the files), Report (PDF)…, Methods text….

Electrode layouts, cleaning, checks, scalp maps, time–frequency, demo expectations, inputs, outputs and troubleshooting are on the EEG page.

Quick start

1. Load data: click Load .mat file. The data type is detected (LDF segments, ERP / average, time series) and the number of series, Fs and time range are shown; the first series is plotted.
2. Parameters: set the stimulus onset t0 (s), the baseline window (s) and the response direction. The plot shows t0 (dashed), the baseline window (shaded) and the detected peak and FWHM, so you can check them before extracting.
3. Features: select the features (Ctrl/Cmd-click for several) and click Extract features.
4. Export: check the table (click a row to plot that series) and click Export to CSV / MAT. Export figure… above the plot saves the selected trace as a publication figure.
5. Groups & statistics tab (or Try group demo): in 1 Files and groups type a Group name and click Add files… (one .mat per animal); repeat for each group. The # column is the subject number: paired designs match #1 with #1, #2 with #2 (fix with ▲ Move up / ▼ Move down).
6. Feature and test: choose the Feature, Value per (File (mean trace) recommended: one animal = one file), Onset t0 (s), Baseline (s) and Direction; then the Design (Paired, Unpaired, ANOVA for 2+ independent groups, or Repeated measures when the same animals appear in 3+ conditions), the Method (Parametric or Nonparametric) and, for two groups, Compare A vs B (difference = B − A). Click Run test: the Plot tab shows every animal (with repeated measures, each animal’s line across the conditions), mean ± SEM (or Box plot) and the significance brackets; the Results tab lists test, statistic, df, p, effect size, 95% CI, n, sphericity and its corrections (repeated measures), a robustness check with the other test family, assumptions and a copy-ready report; the Checks tab lists what was checked before you use the result (n, normality, sphericity, equal spread, agreement of the two test families, missing values; click a row for why it matters and what to try).
7. Export: choose a format and click Export figure… (PDF / SVG / EPS vector, or PNG / TIFF at 300 or 600 dpi; 8.5 cm wide, 8 pt Helvetica, the window is not changed). Export values & report… saves the per-animal values (.csv plus a _report.txt) or the full result (.mat).
8. Session / report (optional): in step 4 of either tab, Save session… stores the single file and every group file (with checksums), all settings and results; Open session… re-extracts the features and re-runs the test; Report (PDF)… writes a one-page summary. See Sessions and reports.

Quick start

1. Pipeline: choose what to run on every file: LDF: trials + response features, LFP: ERP (+ CSD) per channel, MUA: spike sorting per channel, EEG: ERPs and a measure per condition, Imaging: ROI dF/F and vessel diameter or Response features (any trace file). The line below says which files it reads. Try demo batch adds a few synthetic files with known answers and fills the settings.
2. Input files: click Add folder... (every file of the folder that this pipeline reads) or Add files... (multi-select). Select files and click Remove to drop them; Clear empties the list. Files are processed in list order.
3. Settings: one set of settings for every file (hover a field for its unit and meaning). They are the settings of the matching window: e.g. downsampling, filter and trial window for LDF; epoch, N1 window and electrode spacing for LFP; detection, clustering and random seed for MUA; bad channels, filters, reference, events, trial window, rejection, baseline and the measure for EEG.
4. Run: choose the Output folder... and click Run batch. The table and the status bar show which file is being processed; Cancel stops before the next file. A file that fails does not stop the batch: it gets a red row with the reason.
5. Results: one row per file (LFP and MUA: per channel; EEG: per condition; imaging: per ROI). Green = ok, orange = warning (some channels failed), red = error, gray = skipped. Open folder shows the summary (.csv and .mat), the log (.txt) and, for LDF, the trial files; Export... saves a copy of the table (.csv, .xlsx or .mat).

Quick start

1. Save session… (last step card of every analysis window): choose a file name, type optional notes (animal, condition, why these settings) and click Save session. The .nasession.mat file stores the input file paths with their size, date and MD5 checksum, every setting, the results and your notes.
2. Open session…: choose a .nasession.mat saved by the same window. The input files are reloaded and checked; the settings are applied and the analysis is re-run, so the window looks as it did when you saved.
3. If an input file has moved, put it next to the session file (it is found automatically) or choose it when asked. If a file has changed since the session was saved, the session still opens and the status bar warns you.
4. Report (PDF)…: writes one A4 page with a picture of the window and, below it, the Neuronal Data Analyzer Lab and MATLAB versions, the date, every input file with its MD5, the window's checks (warnings first), the settings and the key results. Attach it to your lab notebook or use it for the methods section.