Why each step matters in HVS Image

All distances are worked out from the pool diameter you enter at calibration, so a diameter larger than the real water surface inflates path length and speed. The positions you calibrate are stored by number, and each trial's ‘Platform Number’ refers to them; quadrant measures are only meaningful when platforms sit at quadrant centres.

In the Direct Path implementation, distances are calculated in pixels and converted using the calibrated pixels-per-centimetre value, and the target and close-encounter zones are centred on the calibrated target position. Clicking the pool edges in the wrong order mirrors the plot, so a quick check of one plotted trial against the real platform position protects the whole data set.

Keep all .hvs files: SUM and CST files can be recreated from them. Back up often, and move finished files elsewhere rather than deleting them (moved files no longer show in the list).

Re-analysis matters scientifically: new measures and corrected definitions (for example Direct Path classifications) can be applied to completed studies, and the Direct Path technical note lists retrospective analysis of completed studies as an application. For VR, keep the .BOND files as well as .hvs and .sum.

Each TXP line gives start point, platform, subject and descriptor (for example N,2,Subject1,GroupA), so trial details are entered once, in advance, rather than typed during testing.

Latency is influenced by start point and start-goal distance, so a TXP file that sets out your planned sequence of starts makes that sequence easy to follow, check and report.

Other software can take a fraction of a second or more away from real-time data acquisition, so the tracking (or VR) PC should be dedicated to HVS Image, with scanning and updates run manually outside tracking times.

Symptoms of background interference include the trial clock freezing, dropped frames, briefly lost tracking and straight lines on path plots; each distorts latency, path length and speed, so reliable acquisition is a precondition for valid measures.

Edges are clicked in the order N, S, W, E; then targets, centre first so it becomes position 0. ‘Check Lighting’ and the threshold are set here too, so this one screen controls both scale and tracking quality.

Because target-based measures (crossings, close encounters, Gallagher proximity, Direct Path) are calculated relative to calibrated positions, calibrating every position you may use, including former targets for reversal, lets you analyse those positions later.

Click ‘RUN’ as you release the animal: linking the start to release and detecting arrival objectively avoids premature starts and subjective arrival times. ‘Auto-end on platform’, with an optional delay, gives an objective end point.

End-point choice matters: auto-stop on platform detection counts brief or partial climbs, a dwell delay excludes brief contacts, and manual stop allows experimenter-defined timing. Use the same choice for all groups and report it.

Analyse one trial with ‘Full Analysis’ or many with ‘Analyze Batch or Series’ into a .sum file. ‘Advanced’ settings (counter size, thigmotaxis band, heading distance, time slices) change the results, so record the values you used.

Settings such as Counter diam x and Heading distance change the Direct Path results; report the build and these settings with re-analysed data. Heading range is exported so the value in use is on record.

CST files are comma separated files for Excel and statistics packages, one line per trial. Choosing measures in advance keeps exports consistent across cohorts.

Deciding your primary measures before export, and keeping the same CST definition across cohorts, supports transparent reporting; additional CST files for exploratory measures can be made later from the same .sum.

Results include total path, average speed, zone or cell use, entries and latencies (−1 means not visited), and freezing or inactivity defined by your own time window and movement criteria.

The open field is widely used to assess drug effects on anxiety: Prut and Belzung (2003) found benzodiazepines and 5-HT1A agonists showed anxiolytic effects in most cases, while SSRIs were poorly effective as anxiolytics in the test. Seibenhener and Wooten (2015) describe its use for locomotor and anxiety-like behaviour.

The Detection Area sets how far into an arm counts as an entry: 100% counts any entry, 25% only entries to the far quarter. Per-arm data run from the left-most arm clockwise.

Classic uses include spontaneous alternation in the T-maze, sensitive to hippocampal dysfunction (Deacon and Rawlins, 2006), radial-arm spatial memory (Olton and Samuelson, 1976), and open- versus closed-arm entries in the elevated plus maze as an anxiety measure (Pellow et al., 1985).

The tracking area is a band around the waterline divided into thirds: highest point in the top third counts as climbing, middle third as swimming or floating, bottom third as diving.

The forced swim test was introduced as an animal model sensitive to antidepressant treatments (Porsolt et al., 1977; proposed reference awaiting approval). Interpretation of each percentage depends on your protocol; no forced swim interpretation paper is yet in the verified publications database.

vrScript files set the scene, target, start points and trial types; VR trials are analysed through the water maze Analyse window, so the same measures apply.

The virtual Morris water task revealed severe impairment after hippocampal resection (Astur et al., 2002) and sex differences (Astur et al., 1998). Thornberry et al. (2021) note inconsistent set-ups across more than 200 VR papers and call for standardisation, which is why reporting your vrScript settings matters.

Barnes uses ‘Hole Number’ and ‘Auto-end on escape hole’ in place of the platform captions; the subject size slider is used to ignore faeces.

Barnes (1979) used the circular platform to show poorer spatial memory in senescent rats. It is considered less stressful than the water maze (Rosenfeld and Ferguson, 2014), and water maze training raised corticosterone more than Barnes training (Harrison et al., 2009). Mice can solve it with spatial, cue or serial strategies (Harrison et al., 2006), so strategy analysis is valuable.