Edit this page
Open an issue
Input Directory Setup
This section outlines how the data should be organized prior to running the data2bids software. The data files should be in EDF/EDF+ format following the specifications provided by the EDF format developers.
Setup a working directory is a recommended way to organize your working directory of data (this is optional but highly recommended).
Note
At this moment the names of the EDF/EDF+ files are not yet BIDS compliant, but they do contain some metadata in the filename that will be used later.
Setup a working directory¶
It is recommended that you establish a working directory to ensure your data remains organized. The optimal setup would be one with the following five directories:
- unorganized: this directory holds unorganized subject data ready to be organized for data2bids
- organized: this directory holds organized subject data ready to be organized for data2bids (according to either option below)
- input: this directory holds subject data ready to be converted with data2bids
- output: this directory holds the output converted data from data2bids
- complete: this directory holds the fully converted data from the output directory (which needs to be empty every time a new conversion is occurring)
Warning
When beginning a conversion with data2bids ensure the output directory is empty and the input directory only contains the subject folders you wish to convert.
Static example¶
Interactive example¶
Input directory¶
If you do not require a specific naming convention for visits and sessions then Option 01: Do not specify visit/session number should be followed. In this option, the visit and session number are not supplied to the data2bids software and instead are automatically determined based on the date/time of the EDF/EDF+ acquisition.
For the EpLink study, you should follow Option 02: Specify visit/session number.
Option 01: Do not specify visit/session number¶
If you do not need to specify the visit or session number for each EDF file for the subjects, then this option will assign session numbers based on the Date timestamp within the EDF files. So the earliest EDF file will be given ses-001 while the latest EDF file will be given ses-### (### will be equal to the number of EDF files for that subject).
Input directory
The input directory (input) should contain a sub-directory for each of the subjects you want to have converted (i.e. input\sub-001, input\sub-002 etc.). Within the data2bids software you select input as the input directory and not the individual subject directories
Definition of terms
for a complete list of terms see the definitions page
Static example¶
input/
├── <sub_num>/ # Individual subject directory
│ ├── X~X_432a35cf-adg25-462-24aa-325db4e5e2d3.edf # Individual EDF files
│ └── X~Xe_7d22151a-ac455-3adc312b-426aae3251ac.edf
│
└── <sub_num>/ # Individual subject directory
├── X_ X_35a1ed7a-7764-4cb0-8571-51026e3dbef4.edf # Individual EDF files
└── X_X_e515c5ac-6301-4acd-8a69-fb208d5fd097.edf
Interactive example¶
Option 02: Specify visit/session number¶
In some instances you may want to manually assign the specific visit or session numbers for the EDF files. In this scenario you would need to place each EDF file into a directory with the following naming scheme:
Definition of terms
for a complete list of terms see the definitions page
A folder with the above naming scheme would indicate this is subject 3's second visit and first session. The data collected was a full IEEG recording that was retrospective (recorded prior to the subject consent).
Input directory
The input directory (input) should contain a sub-directory for each of the subjects you want to have converted (i.e. input\sub-001, input\sub-002 etc.). Within the data2bids software you select input as the input directory and not the individual subject directories
Each day of recording should be in a separate folder within the subject directory:
Static example¶
input/
├── <sub_num>/ # Individual subject directory
│ │
│ ├── <sub_num>_<visit_num>_<ses_num>_<type>_<task>_[RET]/ # Specify visit, session, type, and task
│ │ │
│ │ ├── X~X_432a35cf-adg25-462-24aa-325db4e5e2d3.edf # Individual EDF files
│ │ └── X~Xe_7d22151a-ac455-3adc312b-426aae3251ac.edf # Individual EDF files
│ │
│ └── <sub_num>_<visit_num>_<ses_num>_<type>_<task>_[RET]/ # Specify visit, session, type, and task
│ │
│ └── X~X_432a35cf-adg25-462-24aa-325db4e5e2d3.edf # Individual EDF files
│
└── <sub_num>/ # Individual subject directory
│
└── <sub_num>_<visit_num>_<ses_num>_<type>_<task>_[RET]/ # Specify visit, session, type, and task
│
├── X~X_432a35cf-adg25-462-24aa-325db4e5e2d3.edf # Individual EDF files
└── LastName~ FirstName_7d22151a-ac455-3adc312b-426aae3251ac.edf # You can include the subject first/last name to be used when de-identifying the data
Interactive example¶
Channel labels¶
For recordings where patient specific channel labels are defined, a channel labels text file should be created that dictates how the labels should appear within the EDF file. For instance, EEG recordings commonly make use of the 10-20 setup, which means the channel labels do not change between patients. However, in iEEG recordings, channels labels are unique to the patient depending on where the depth electrodes were implanted. To ensure the EDF file has the correct channel labels, you should create a text file called "*_channel_labels.txt", where * indicates any string can be placed there but the file name needs to end with channel_labels.txt (i.e. sub-001_channel_labels.txt, EPL31_LHS_0001_channel_labels.txt).
When creating the channel labels file you do not need to include default NeuroWorks channels, such as:
- Patient Event
- TRIG
- OSAT
- PR
- Pleth
- EDF Annotations
You only need to provide channels with unique names for the specific patient, for instance:
- LAHc1
- RAm1
The implanted depth electrodes each have several contacts on them to record from. So the contacts on the electrode are numbered, such as:
- 10 contact electrode LAHc: LAHc1, LAHc2, LAHc3, LAHc4, LAHc5, LAHc6, LAHc7, LAHc8, LAHc9, LAHc10
When creating the channel labels file, it is easiest to use Excel or Libreoffice. The first column should contain C1 to Cn, where n is the total number of iEEG channels used in the recording. Next to each channel number, place the corresponding electrode contacts name.
Example _channel_labels.txt file