Time-sync neural recordings with inertial sensor recordings

Hi There,

I am recording neural activity with implanted electrodes onto a pacemaker implanted in a patients chest. The wire for the electrodes run under the skin of the neck. I also will record motor control with inertial sensors (magnetometers, accelerometers and gyroscopes). I need to time-sync these devices. The neural recording will be running as well as the inertial sensor software. I need a way to define "time-zero" so that I can analyze the data offline.

I wanted to use an Arduino to send a pulse to an EMG pad on the patient's neck (just above the wire) in order to introduce an artifact in the neural recording. At the same time I wanted to also send a trigger signal to the laptop running the inertial sensor software to trigger a marker in the recording. In this way I will have marked a "time-zero" on both devices.

How would I go about doing this?

Thank you,

Greydon

Hello Greydon,

I guess you would like to measure neural reactions of someone sat on a vehicle accordingly with the vehicle position, you would need recordings of neural reactions timed with the vehicle positions, i.e. time measurement should obviously be consistent between the two recordings.

The 32 bits ARM Sam3x8e embedded on the Arduino DUE board offers a Real Time Timer (RTT), a Real Time Clock (RTC) and a bunch of timers.

You can trigger recordings in a starting window of a few nano seconds with Systick, a few micro seconds with micros() or any Timer or simply use RTT or RTC for a precision of a second.

If the starting point should be very precise then you don't want latency in your recordings, i.e. you will certainly use a DMA memory to memory or peripheral to memory.

To handle a practical implementation you'd need to provide some buffering, either in the form of a ring buffer, or a collection of buffers, which would be feed by the DMA unit.

You need to worry about the latency of the laptop recording - can you induce a direct
signal on the IMU signals simultaneous to the EMG pad signal? perhaps a magnetic pulse for
the magnetometer, or a little tap from a small solenoid that would show up on the accelerometer
signal?

This is a clapperboard in effect - something to be observed by two different media recording
devices