Py-OCT-Calibration

This project aims to produce an easy to use tool to calibrate any SD or SS-OCT. The processing is based on Attendu et al. "Simple and robust calibration procedure for k-linearization and dispersion compensation in optical coherence tomography." (doi: 10.1117/1.JBO.24.5.056001)

By calibrate I mean substracting background - k-linearize spectra - evaluate and compensate dispersion. In order to compute the calibration the algorithm need multiple measurements:

mirror1.npy (OCT interferogram with mirror in one side of the zero delay as sample)

dark_sample.npy (OCT interferogram with ref arm blocked)

mirror2.npy (OCT interferogram with mirror in the other side of the zero delay as sample)

dark_ref.npy (OCT interferogram with sample arm blocked)

dark_not.npy (OCT interferogram with both arm blocked)

Those .npy file should be saved into a file in data/calibration/<your-name> directory.

Packages dependencies

In order to use the Py-OCT-Calibration Library, one must have installed the following packages:

    - Numpy
    - Scipy
    - json
    - matplotlib
    - Cupy (fot Nvidia GPU acceleration)
    - sphynx (for developer only)
    - git (for developer only)
    - Napari (for 3D-rendering)
    - profile-viewer (for developer only)

To install CuPy librarie, one should refer to their website, link: CuPy installation

For the other dependencies and using pip3 one can use the following commands:

>>> pip3 install Numpy
>>> pip3 install Scipy
>>> pip3 install matplotlib
>>> pip3 install gitpython
>>> pip3 install Cupy
>>> pip3 install napari
>>> apt-get install python-sphinx (for Unix OS)
>>> sudo port install py27-sphinx (for Mac OS)

Project architecture

The folder architecture is presented as :

    - setup.py                          (setup installation tool for compilation)

    - makefile                          (for tests and installation)

    - data/                             (contain all data)
        - calibration/
        - Aline/
        - Bscan/
        - Cscan/

    - src/
        - processing/                
            - calibration.py            (compute calibration parameters)
            - Aline.py                  (process one Aline)
            - Bscan.py                  (process one Bscan)
            - Cscan.py                  (process one Cscan)
            - post_processing.py        (3D rendering of Cscan volume)

       - toolbox/                      (contain tools to do all the processing)

       - Doc/                           (not yet fully developped)
            - images/                   (images for README.md)
            - profiling/                (for optimization endings)
            - documentation/            (auto-generated documentation)           

    - README.md                         (literally this)

Python init.py and compiled files are not represented here.

Run tests

Verify installation is correct by running some tests:

>>> make test_calibration
>>> make test_Cscan_cpu
>>> make test_Cscan_gpu      (only with CUDA installed)
>>> make build_doc           (generating html documentation)

Run example

Calibration

In order to run a calibration example one can tape the following command on command prompt:

>>> python3 src/processing/calibration.py --dispersion=1 --input-dir=./data/calibration/example/ --output-file=data/calibration/example/calib.json

Arguments for src/processing/calibration.py are:

--dispersion : [1] for normal dispersion, [-1] for anormal

--input-dir : directory of the input files for calibration

--output-file : name of the output .json file containing all the calibration parameters which is saved in same directory as input-dir

Denoising process:

Phase computing:

k-linearization/disperison compensation:

Aline

In order to process one example of Aline, one can tape the following command on command prompt:

>>> python3 src/processing/Aline.py --dispersion=1 --input-file=data/Aline/example.txt --calibration-file=data/calibration/example/calib.json

Arguments for src/processing/Aline.py are:

--dispersion : [1] for normal dispersion, [-1] for anormal

--input-file : directory of the input Aline file

--calibration : directory for the output .json file containing all the calibration parameters

--output-file : directory for the processed Aline file

Bscan

In order to process one example of Bscan, one can tape the following command on command prompt:

>>> python3 src/processing/Bscan.py --dispersion=1 --input-file=data/Bscan/example.npy --calibration-file=data/calibration/example/calib.json

Arguments for src/processing/Bscan.py are:

--dispersion : [1] for normal dispersion, [-1] for anormal

--input-file : directory of the input Bscan file

--calibration : directory for the output .json file containing all the calibration parameters

--output-file : directory for the processed Bscan file

Cscan

In order to process one example of Cscan, one can tape the following command on command prompt:

>>> python3 src/processing/Cscan.py --dispersion=1 --input-dir=Cscan/example/ --calibration-file=data/calibration/example/calib.json --dimension 100 100 1024

Arguments for src/processing/Cscan.py are:

--dispersion : [1] for normal dispersion, [-1] for anormal

--input-file : directory of the input Aline file

--calibration : directory for the output .json file containing all the calibration parameters

--output-file : directory for the processed Cscan file

Documentation

Documentation on project is available in a html file "docs/documentation/build/index.html"

For citing this project in scientific papers

Pep8 coding convention

In order to keep a clean and consistent code, one can follow the convention as presented in the following link:

https://www.python.org/dev/peps/pep-0008/#documentation-strings

MartinPdeS/OCT_calibration