Tutorial: Analog Output Control

This lesson starts out from the default setup found in ttDefault.mat.

Analog Control - Stretch Reflex

A wide variety of equipment may be controlled by an analog signal: An example could be an actuator used to move and/or position a device, e.g. a pedal or alike device to induce stretch reflexes.

The measurement of stretch reflexes will be used as an example below: The Soleus muscle will be stretched by a ramp-hold-ramp rotation of the ankle joint, i.e. a pedal will hold the ankle joint at a position, then rotate it at a specified rate to another position, stay there for a short while and finally rotate back to the original steady position. The stretch of the muscle will elicit a reflex, which appear in both the EMG and the force output from the muscle.

Stretch Reflex, Simple

First we'll take a look at a very simple example: We'll acquire a series of equal stretches. The setup derived in the following can be found in ttAOstr0.mat.

Acquisition

The stretch reflex can be seen in the EMG approx. 40 to 200 ms after stretch onset. The resulting force output from the muscle will peak 200-300 ms after stretch onset and last for several hundred ms. A total sweep length of 800 ms should be sufficient and we'll choose 200 ms pre-trig to monitor activity just before the stretch.

A SampleRate of 2 kHz is reasonable for the measurement of EMG, whereas force and position may be sampled at 500 Hz - a factor of four lower than the EMG.

Analog Input

For this purpose we'll need only one EMG and one force (or torque) channel. It may still be reasonable to monitor more channels: We'll choose four muscles (EMGs) along with both force and position. Each channel is labeled according to the signal measured. All EMGs will be sampled at high sample rate, rectified and low-pass filtered. Force / torque and position are sampled at low sample rate and grouped as Kinematic.

All sources supply voltages in the range from -10 to +10 V.

Classification: Y-Analysis

Even though Classification is not in play in this simple example we may still use the Display Y-Analysis panel to give us an idea of the size and variation of the stretch reflex during acquisition. Let's monitor the Maximum peak in the Soleus EMG in the time interval from 40 to 140 ms after stretch onset.

You can show/hide the Y-analysis display via the Windows menu. It will automatically hide if no classification is applied.

Trigger

We want to induce stretches and acquire reflexes at a regular interval, say approx. every four seconds. We'll choose Internal as the trig Mode/Source for Mr. Kick to control this. In order to avoid anticipation stretches should not be applied exactly every four seconds. By setting Refractory to 3.5 s and Max Interval to 4.5 s the interval between stretches will vary randomly ±0.5 s around a 4 s mean.

Event Timing

In this example no external equipment is trigged by the event timers. Only data acquisition is trigged by event timer #0. We'll set Delay to zero and Duration to 10 ms, Pulse Polarity High.

Analog Output

One analog output channel is needed to control the stretch device: We choose First AO Channel setting Active to Primary (no superimposition of different patterns are needed for this). The Sensitivity is set to reflect the stretch device in use - let's assume the pedal will move 0.5 degree for each volt supplied to its controller.

In the D/A section check that the AO trig Source is synchronized to Timer 0 / AD and the Update Rate is reasonable, say 1 kHz.

It's now time to load the plug-in that should generate the analog output pattern: In this case this will be the AOPI Ramp-Hold-Ramp.vi. The Ramp-Hold-Ramp SetUp will pop up to allow you to specify the type of stretch you want. We'll go for a 4 degree stretch Amplitude with a Rise Time and Fall Time of 40 ms and a Hold Time of 460 ms (Duration 500 ms). It should not be repeated and therefore the Interval becomes irrelevant.

Stretch Reflex Dependency on Velocity

It may be interesting to investigate how the size of the reflex depends on stretch amplitude, velocity and alike. In the following we'll extend the above so that a range of stretch velocities can be investigated in one experiment. The major advantage of this is that the different stretch velocities will be applied at random. This will ensure that the results for different velocities are achieved under comparable circumstances and thereby improve the credibility of the outcome of the experiment.

The setup derived in the following can be found in ttAOstr1.mat.

Acquisition, Analog Input, Y-Analysis, Trigger and Event Timing

Same as above.

Classification: Stimulation

In order to control stretch parameters (stimuli) we set Classification mode to Stimulus Type. And we'll let classification Act On Analog Output, which includes the rise time parameter that controls stretch velocity.

The Nmb Classes should be set to 8 in order to implement the eight different rise times given below. Mr. Kick are asked to pick the different velocities (stimulation classes) at random.

Analog Output

The stretch velocity is controlled through the rise time of the ramp-hold-ramp pattern. For the eight (sub) classes made available above we'll set Rise Time to the following values: 20, 40, 60, 80, 120, 160, 250, 320 ms. Hold time may be reduced accordingly: 480, 460, 440, 420, 380, 340, 250, 180 ms.