General Parameters

If a trial's weight is 0, it will not be presented at all. In all other sequencing modes, the weight is ignored. On the Trial Editor, the weight is displayed and modified by a spinner control with editable "buddy" window. Change the weight's value by entering a valid integer in the edit window, or by clicking or holding down the spinner's up and down arrows. Allowed range is [0..255].

Recording and displaying trial data

The trial designer can choose whether or not to save a given trial's data, specify when recording should begin, and select which data channels are saved to file and which are displayed during runtime.

The KEEP/TOSS pushbutton indicates whether the trial's data is saved or discarded. Simply click the button to toggle its state. In practice, this is almost always set to KEEP, which is the default value.

Often it is undesirable to record the first portion of a trial, during which the animal is given some time to fixate on a central point prior to presenting the test protocol. The spinner labelled Starting at seg determines when Maestro begins recording a trial's data. Recording is turned on at the beginning of the segment displayed in the spinner's buddy window. If the segment index is -1 or 0, the entire trial is recorded. Click on the spinner's up and down arrows to increment or decrement the segment index. [Note that Maestro automatically adjusts this index -- and all other trial parameters that are set to a segment index-- as the user inserts or removes segments from the trial.]

When a trial aborts because the subject violated fixation requirements, Maestro will normally discard the data file. However, if the trial runtime reached the beginning of the "failsafe" segment, the data file will be saved regardless. Use the Failsafe seg spinner to set the index of the failsafe segment. If the segment index is -1 (the default), then the trial must run to completion. If the trial aborts because of a runtime error such as a video frame drop, the data file is always discarded, even if a failsafe segment is specified.

The dropdown combo box labelled Channels selects the channel configuration associated with the trial. The channel configuration determines which of Maestro's 16 analog channels are recorded and saved to file during the trial -- typically only a small subset of the available channels are actually saved. In addition, it selects up to 10 data channels (analog, digital, or computed target trajectory) for display in the Data Traces display panel and sets the gain, offset, and color of each data trace. The combo box's dropdown list will include all channel configuration objects defined in the current experiment document, plus a "NONE" entry. This last entry is rarely used, since no analog data will be saved and nothing displayed during the trial.

When a trial is longer than a few seconds, it is difficult to see much detail in the data displayed in the Data Traces panel. In this case, the user can choose to display only a portion of the trial spanning a contiguous range of segments. The endpoints of this segment range are set by the two spinner controls following the Display Interval label. Given segment indices 0 <= segA <= segB, the display interval [t0..t1] will encompass the segment range [segA..segB]. If either segment index is -1, then the entire trial will be displayed.

Reward pulses and random reward withholding

Rewarding the animal when it successfully completes a trial is obviously fundamental to any experiment, and trials in Maestro offer considerable flexibility with reward delivery: an adjustable reward pulse length to set the "size" of the reward; two separate rewards for two-choice experiments, and the ability to randomly withhold an earned reward.

All experiment rigs must be equipped with an "adjustable reward pulse" device. This auxiliary device, part of a comprehensive Digital I/O interface and controlled by the 16-bit digital output port of Maestro's DIO event timer board, generates a pulse of adjustable duration. This pulse gates the solenoid that delivers a liquid reward to the animal.

Two reward pulses. Maestro trials offer two independent reward pulses. A typical trial delivers reward pulse #1 upon successful completion of the trial, and reward pulse #2 is unused. However, the second reward pulse comes into play in a number of scenarios:

• • In trials featuring a selByFix special operation, reward pulse #1 is delivered at trial's end when the animal "selects" the first to two fixation targets, while reward pulse #2 is delivered at trial's end when the second fixation target is selected.

  • In trials featuring a chooseFix special operation, reward pulse #1 is the end-of-trial reward as usual, while reward pulse #2 governs the duration of the reward delivered if and when the animal fixates (ie, its eye position is within the fixation window of) on the "correct" target during the special segment.

  • In trials featuring a R/P Distro operation, reward pulse delivery depends upon whether or not the "response reward window" has been enabled for the trial. The reward window is not part of the trial's definition; rather, it is defined in the R/P Distro tab of the Trial mode control panel. If the reward window is disabled, then reward pulse #1 is delivered at the end of the trial in the usual manner. If it is enabled, one or both reward pulses may be delivered. If the subject's measured behavioral response falls within the window, reward pulse #2 is delivered at the end of the special segment AND reward pulse #1 is delivered at the end of the trial -- assuming the trial is completed normally. If the subject's response is outside the reward window bounds, then the second reward pulse is delivered only at the end of the trial. By making the second pulse shorter than the first, the experimenter ensures the animal will get an enhanced reward when its response is within a reward window, and a reduced reward otherwise.

The selByFix, chooseFix, and R/P Distro features are discussed in greater detail on the Special Features page.

Random reward withholding (as of Maestro v4.1.0). In behavioral paradigms, continuous reinforcement of the desired behavior during initial training is often followed by a partial reinforcement schedule, in which the subject does not get the reinforcement (aka, reward) every time they perform the behavior. A common type of partial reinforcement is "variable ratio" N/D, in which the reward is withheld for a random selection of N correct responses out of every D repetitions of the behavioral task. Maestro supports per-trial specification of the withholding variable ratio (WHVR) for each of the two possible reward pulses.

As shown in the above screen capture of the General parameters group, each reward pulse is governed by three numeric edit controls: the pulse length P, the WHVR numerator N, and the WHVR denominator D. Pulse length P is specified in milliseconds and is restricted to integer values in the range [1..999], with a default value of 10. The WHVR numerator and denominator are integer values that must satisfy 0 <= N < D <= 100; the default values are N=0 and D=1, which disables random reward withholding.

[NOTE that Maestro also features a global WHVR defined in the application settings and exposed on the Fix/Reward control panel. The global and per-trial WHVR features are incompatible; if you use per-trial WHVR, be sure the global WHVR is disabled. Also, when the per-trial WHVR feature is in use, analysis programs can easily detect a trial in which the reward was earned but randomly withheld due to per-trial WHVR. The flag bits in the data file header will include the CXHF_REWARDEARNED flag but not the CXHF_REWARDGIVEN flag; in addition, the reward pulse length in the header field iRewLen1 or iRewLen2 will be zero if the reward was indeed withheld.]

Overriding global target position and velocity vector transforms

On the Other Params tab of the Trial mode control panel, the user can specify transforms for scaling and rotating the position and velocity vectors of each target participating in a trial. The transforms are applied to every target and on every segment of the trial (with one exception: in segment 0 only, target transformations are not applied to any target that is positioned ABSolutely). These transforms are most often used to tailor a set of trials to the directional selectivity and/or speed sensitivity of a particular neural unit during the experiment -- without having to modify each individual trial. In some testing scenarios it may be useful to apply only a subset of these transformations to some trials in a trial set, but not all of them. The four checkboxes to the left of the label Ignore transforms let the user choose which of the global transformations should be applied to the trial's target trajectories. Check the appropriate box to disable the corresponding transform. If you do not use the global transform feature in your work (that is, the scale factors are 1 and the rotation angles are zero), it is best to leave all boxes unchecked.

[NOTE: The velocity vector transforms are also applied to target window and pattern acceleration vectors.]