Hi all,
I’m currently doing an honours project which involves hopefully controlling the AL5A arm (serial interface with the SSC-32 card) using only facial EMG and EOG physiological commands. We normally use ADI Instruments’ Powerlab, and Matlab with these physiological signals. We’ve used a serial monitor to log and understand the codes that the computer is sending via serial port to the arm for the different movements, so now we’re hoping to to create a programming circuit whereby the computer will send commands to the arm that are proportional to the changes in voltage in the EOG/EMG signal (basic example: look left = rotate arm left). We’re just up to the stage where we’re trying to determine how best to read and use the voltage changes from the electrodes.
Has anyone here used Powerlab and/or Matlab before to program serial devices with physiological signals, or does anyone have any other suggestions?
Thanks
First, you might define EOG/EMG signals.
What levels are you getting, what interface do you have to Powerlab and Matlab?
Alan KM6VV
We’ll firstly focus on EOG (Elecroocculogram) signals, which are changes in electrical activity based on eye movement. Since the eyes are polarised, by placing electrodes on the outer sides of each eye we able to record these differences. For example, look left might induce a positive voltage in the left electrode, and thus a negative voltage in the right electrode, and vice versa.
These changes are obviously pretty small, so we’ll be using a preamp to amplify the signals, which I think Powerlab has. I’m not sure on the level of the signals, or what it needs to be though yet.
I don’t know too much about Powerlab and Matlab myself, since I have hardly used it before and am relying more so on my colleagues’ knowledge of it, however it seems Powerlab is essentially the hardware interface for the electrical signal from the electrodes into the computer, which is then analysed using the Matlab software.
It seems it’s possible to use these electrical signals, combined with the serial commands we already know, to control the arm. Additionally, Powerlab appears to have an output port…
There have been some previous studies which have used eye movement or muscle movement to control robotics, so I’m just wondering if someone here has done something similar, and how they did it? Any bit of advice will help and if someone has done something similar using Powerlab or Matlab, we’d be interested to hear the set-up used.
From what I’ve seen, serial port operations with matlab/labview is just about what you can do with a batch file. For the $$$… paid for these programs, the vendor support appears to be pitiful.