File and directory naming conventions for Bruker NMR data
Bruker data is classified according to the following conventions:
- Each individual specimen is considered a Patient and is given a Patient Name and unique Patient Registration ID (10 alphanumeric characters long). We try to keep to this convention of one specimen = one Patient, but sometimes if the number of specimens is large then we group several subjects into the same Patient.
- Each Patient can contain several Studies, each of which ostensibly correspond to one scan session. Each Study is assigned a Study Name (which the user inputs) and Study ID. The Study ID is assigned automatically by the scanner software, and corresponds to the nth study created on the current day for the current Patient.
- Each Study contains several Experiments, which correspond to one MRI/NMR scan acquisition. Each of these scans are assigned a unique ExpNo.
- Each Expno can contain several Processed Datasets. Usually, there is only one Processed Dataset which contains the image set, but there may be others if special image processing was required. Each Processed Dataset is assigned a ProcNo.
- Each Study contains several Experiments, which correspond to one MRI/NMR scan acquisition. Each of these scans are assigned a unique ExpNo.
- Each Patient can contain several Studies, each of which ostensibly correspond to one scan session. Each Study is assigned a Study Name (which the user inputs) and Study ID. The Study ID is assigned automatically by the scanner software, and corresponds to the nth study created on the current day for the current Patient.
This naming convention influences the way the data directories and files are organized:
Bruker software creates a data directory for each Study of a Patient named
<Patient_Registration_ID>.
where xx is a 2-character code for the date that the Study was created (not easily decipherable) and y is the Study ID. Within this directory, there is the subject text file, which contains the Patient and Study-specific parameters (e.g. Patient and Study Name, Study creation timestamp, etc.). The directory also contains further subdirectories which contain the individual Experiments, named after their ExpNo (i.e. 1, 2, 3, etc.). Each Experiment directory contains another subdirectory called pdata as well as several data files:
- fid: the raw unreconstructed data file (i.e. the “k-space” time-domain data)
- method: text file containing the high-level acquisition parameters (e.g. number of pixels, measurement method, field of view, echo time, repetition time, number of averages, etc.)
- acqp: text file containing the base-level acquisition parameters (the method parameters are derived from these acqp parameters, which are a bit like the “machine-code” variables of the pulse sequence that was used)
- pulseprogram: text file containing the MRI sequence
The pdata folder only exists for image data. It contains subdirectories corresponding to the different Processed Datasets of the current Experiment – each directory is named after the (e.g. 1, 2, etc.). For the majority of cases, the “1” directory contains the processed image data. Files in “1” include the following:
- 2dseq: the processed image data, expressed in a raw binary format without header. If there are multiple slices in the dataset, they are usually stored contiguously. The default binary word type is 16 bit signed integer, little endian.
- Reco: text file including all reconstruction parameters
Sometimes you will find some of these directories or files missing. This is likely because the Experiment was created but never started (i.e. no data was acquired).
BOTTOM LINE: Most users will only need to know how to find the 2dseq file (for image data), as well as the method and reco files for text information about the acquisition and reconstruction parameters.