Micromed IO library

A library to read, emulate, and forward Micromed data in standard formats. See online doc.

Main features:

simulate online data from a trc file
push online tcp data to LSL server
convert trc to mne format
rename trc files to include the recording datetime

Install

$ pip install micromed-io

Convert a Micromed (.trc) file to MNE (.fif) format

from micromed_io.to_mne import create_mne_from_micromed_recording
mne_raw = create_mne_from_micromed_recording("path/to/file.TRC")

Emulate TRC to TCP & read/forward to LSL server

See details in next sections

Emulate Online Micromed TCP from .trc file

$ mmio_emulate_trc --file=../data/sample.TRC --address=localhost --port=5123

Emulate the online data stream of Micromed to test your real-time platform. See all the arguments and adapt them:

$ mmio_emulate_trc --help # to see all arguments

Note

New: not only data, but also markers and notes are send through TCP, exactly as Micromed does

Read TCP and push to LSL Stream

$ mmio_tcp_to_lsl --address=localhost --port=5123

While receiving online data throug tcp, this command forward the data to 3 LSL stream outlets:

Micromed_EEG: the eeg data in float32 format [n_channels, n_samples]
Micromed_Markers: markers if any in int32 format [sample, marker] (2 channels)
Micromed_Notes: notes if any in string format [sample, note] (2 channels)

You can easily change the LSL parameters:

$ mmio_tcp_to_lsl --help # to see all arguments

Read TRC file

from micromed_io.trc import MicromedTRC
mmtrc = MicromedTRC("sample.TRC")

Then you have access to the trc data:

mmtrc.get_header()
mmtrc.get_markers()
mmtrc.get_data()
mmtrc.get_notes()

Note

Note: get_data() might take times because it loads the brain data

Read and parse Micromed TCP live data

Download tcp_to_lsl.py from the github repo in scripts/

$ python tcp_to_lsl.py --address=localhost --port=5123

Note: Micromed TCP behaves as a client. If you want to try the emulate/read TCP script, launch the reader first that acts as server, then the emulator.

Rename TRC files with recording datetime

$ mmio_rename_trc --dirpath=./ --format=%Y%m%d-%H%M%S

Rename the TRC files of the given folder to include the recording date in the filename. Output is : <filename>__<recording_date>.TRC. The format must be compliant with python strftime format codes

mmio_rename_trc --help # to see help

Local install

Download the repo and:

$ conda env create -f environment.yml
$ conda activate mmio
$ poetry install

Please feel free to reach me if you want to contribute.

Sponsor

This work has been supported by the Wyss Center.

etiennedemontalivet/micromed-io