ORPL (read orpel) is the Open Raman Processing Library. It provides tools for the processing of Raman spectrum, including;
- System calibration (x-axis and system response)
- Cosmic Ray removal
- Baseline removal (autofluorescence)
- Spectrum analysis (peak finding, ...)
- Synthetic spectrum generation (for testing and benchmarking)
As of v1.0.0, ORPL also provides a Graphical User Interface. See demo below ;)
I was notified of a few compatibility problems with the use of ORPL on non-linux systems. Most of the difficulties were traced back to making numba behave with MacOS and Windows. Starting with version v1.0.10a, I removed numba
from the required libraries list. As such, it will no longer be installed on its own along orpl
. However, if it is installed explicitely (by running pip install numba
), it will provide a significant boost for the performance of baseline removal filters. In the future I plan to find another library to handle JIT compilation to achieve the same benefits.
In addition, I was also notified of an error that a few people encountered when trying load a spectrum file in ORPL GUI. The GUI would launch fine, but would crash when trying to load a spectrum. Part of the error message is as follow,
"dump()" has been removed, use
yaml = YAML(typ='unsafe', pure=True)
yaml.dump(...)
This error is caused by a newer version that was pushed for the ruamel library that I use to parse YAML. I will need to change a few things, but for now an easy fix was to limit this dependency to versions prior to 0.18. If you encountered this error, there are two ways to fix it.
First, you can explicitely install ruamel with a version that does not cause this issue,
pip install "ruamel.yaml<0.18.0"
Or you can force an orpl
update,
pip install --upgrade orplib
- ORPL
orpl_GUI_demo_1.mp4
At its core, ORPL is designed to be a processing library to use in your own processing workflow. Nevertheless, I also wrote a GUI to go with it if programming is not your jam. In either case, installation is made through pip.
I wrote a detailed installation guide for windows complete with screenshots of all the steps, so do not worry if you are a python beginer and have no idea what I'm talking about here. You can access the guide on github or download a version.
I recommend you create a virtual environment with venv. Otherwise, just install orplib with pip.
pip install orplib
Using Anaconda?... dont... Jokes aside, if people ask me about this, I might write a guide for this. Otherwise, use pip.
I am working on a python tutorial repository, you can learn more about it .
If you want to update to the latest version of ORPL, run the following pip command,
pip install --upgrade orplib
pip install --upgrade --user orplib
This is the command you need to run if you want to build the .whl from the source code yourself (make sure you run it from orpl's project root directory):
python -m build
and to update the build on pypi (this is a reminder for me, it won't do anything if you do this),
python -m twine upload --repository pypi --skip-existing dist/*
You can download sample files to play around with ORPL - GUI or the library. They are located here
Bubblefill is a morphological processing algorithm designed for the removal of baselines in spectroscopic signal. It was created and optimized specifically to remove autofluorescence baselines in Raman spectra measured on biological samples.
The tuning parameter of Bubblefill is the size of the smallest bubble allowed to grow. In general, the smallest bubble width should be chosen to be larger than the widest Raman peak present in the signal. Otherwise, the baseline fit will grow inside the peaks and the output Raman signal will have under expressed peaks.
Note : Bubbles can become arbitrarily small if they are growing along the leftmost or rightmost edge of the signal.
Different smallest bubble widths can be specified for different regions of the spectrum. This enables nearly infinite tuning of the algorithm and can be used to remove peaks that are known artifacts (for instance). In this example, the smallest bubble width for detector pixels 400 to 650 was set to 1 and to 100 for the rest of the x-axis.
Guillaume Sheehy, Fabien Picot, Frédérick Dallaire, Katherine Ember, Tien Nguyen, Kevin Petrecca, Dominique Trudel, and Frédéric Leblond "Open-sourced Raman spectroscopy data processing package implementing a baseline removal algorithm validated from multiple datasets acquired in human tissue and biofluids," Journal of Biomedical Optics 28(2), 025002 (21 February 2023). https://doi.org/10.1117/1.JBO.28.2.025002
@article{10.1117/1.JBO.28.2.025002,
author = {Guillaume Sheehy and Fabien Picot and Fr{\'e}d{\'e}rick Dallaire and Katherine Ember and Tien Nguyen and Kevin Petrecca and Dominique Trudel and Fr{\'e}d{\'e}ric Leblond},
title = {{Open-sourced Raman spectroscopy data processing package implementing a baseline removal algorithm validated from multiple datasets acquired in human tissue and biofluids}},
volume = {28},
journal = {Journal of Biomedical Optics},
number = {2},
publisher = {SPIE},
pages = {025002},
keywords = {Raman spectroscopy, fluorescence, tissue optics, open-sourced software, machine learning, optics, Raman spectroscopy, Data processing, Bubbles, Equipment, Tissues, Biological samples, Raman scattering, Fluorescence, Aluminum, Spectroscopy},
year = {2023},
doi = {10.1117/1.JBO.28.2.025002},
URL = {https://doi.org/10.1117/1.JBO.28.2.025002}
}
%0 Journal Article
%A Sheehy, Guillaume
%A Picot, Fabien
%A Dallaire, Frédérick
%A Ember, Katherine
%A Nguyen, Tien
%A Petrecca, Kevin
%A Trudel, Dominique
%A Leblond, Frédéric
%T Open-sourced Raman spectroscopy data processing package implementing a baseline removal algorithm validated from multiple datasets acquired in human tissue and biofluids
%V 28
%J Journal of Biomedical Optics
%N 2
%P 025002
%D 2023
%U https://doi.org/10.1117/1.JBO.28.2.025002
%DOI 10.1117/1.JBO.28.2.025002
%I SPIE
- Guillaume Sheehy | [email protected]
- Frédérick Dallaire
- Fabien Picot