Naming of motion fields

[Rob Luke]

Hi all,

In BEP030:fNIRS we included channel types for accelerometers, gyroscopes/gyrometer, and magnetometers, as these are commonly included in off-the-shelf devices and are stored in the same file as raw fNIRS data.

However, these motion channels are common across various devices, not just fNIRS (for example, some EEG devices include these measurements). Additionally, there are specific motion devices, and BEP029:Motion is being developed to track a wide range of motion capture systems used in human research, including accelerometers, gyroscopes, and magnetometers.

Consequently, we would like to engage the wider community in the naming of these types. It does not seem appropriate for the fNIRS BEP to decide the naming of these channels alone and force all other modalities to follow suit.

In BEP030 we propose adding types ACCEL, GYRO, MAGN.

In BEP029 they propose adding ACCEL, ANGVEL, MAGN.

Everything is quite consistent except the naming of GYRO vs ANGVEL (both referring to the same thing, data from a gyroscope).

One argument for using GYRO (copied from Robert Oostenvelds comment) is to "name channels to the type of measurement, not to the value that is represented in the channel. For EEG we also don't call it a VOLT channel.". Further, I would argue that most people refer to these measurements as gyro (although this may be specific to the field I work in).

And from Sein Jeung "my concern about naming gyrometer output "GYRO" is that this logic does not apply to any other motion data type such as positions, or joint angles.

I would like to request the input of the wider community on the naming of these channel types (ACC, GYRO/ANGVEL, MAGN). Both your input on the general naming of the three types, and specifically the naming of GYRO vs ANGVEL.

Thank you in advance,

Rob

[julius welzel]

Dear Rob,
thanks for bringing this to a broader community. I am used to working with IMUs and we usually name channels according to the devices we work with (ACCEL for Accelerometer, GYRO for Gyroscope and MAGN for Magnetometer). I think these are the most common names in our community. In my opinion, the type should be easy to understand and does not need to follow strict rules. I still agree with Sein that this can be confusing for people who are not familiar when working with motion data and neither Gyro nor AngVel is commonly used.
Best and a happy weekend to everyone,
Julius

[Jörn M. Horschig]

Hi,
I agree with  Robert and Julius that the terms accelerometer, gyroscope and magnetometer are more commonly used. Regarding Sein Jeung's comment, I see his point that position and joint angles are other "channel" names that can be used, but then ACCEL and MAGN should be renamed as well. Since consistency is king, I'd name it GYRO.

Best,

Jörn

[Sein Jeung]

Hi everyone, sorry that I saw this thread a little late. 

I'd be fine with going with GYRO label at this point, although the label GYRO may stand out a bit when saved with other motion data types. 
From the discussion I understood that Gyro is what people are used to when listing IMU components.
Besides I don't expect many scenarios when people share angular velocity data that is NOT measured from a gyroscope.
If someone really derives angular velocity from orientation or some other data, it would then fall out of scope already because that wouldn't be minimal preprocessing. 

Best, 
Sein