From 229ed6c00dda12cea828d9275237ecfbfe1ca120 Mon Sep 17 00:00:00 2001 From: MichaelSNelson Date: Mon, 7 Oct 2024 00:12:32 +0000 Subject: [PATCH] deploy: e7eb64c06638f063d7d9b0f3a082f0a58cd14633 --- .../1_established_workflows.md | 28 +++++--- .../analysis_workflows/2_new_workflows.md | 53 +++++++++++---- .../3_machine_learning_workflows.md | 66 ++++++++++++++----- _sources/image_publishing/1_image_format.md | 53 +++++++++++++-- .../2_image_colors_channels.md | 36 ++++++++-- .../image_publishing/3_image_annotation.md | 4 +- .../1_established_workflows.html | 23 ++++--- analysis_workflows/2_new_workflows.html | 41 ++++++++---- .../3_machine_learning_workflows.html | 54 ++++++++++----- image_publishing/1_image_format.html | 41 ++++++++++-- image_publishing/2_image_colors_channels.html | 31 ++++++--- image_publishing/3_image_annotation.html | 4 +- searchindex.js | 2 +- 13 files changed, 333 insertions(+), 103 deletions(-) diff --git a/_sources/analysis_workflows/1_established_workflows.md b/_sources/analysis_workflows/1_established_workflows.md index c3a806a..19fdbd3 100644 --- a/_sources/analysis_workflows/1_established_workflows.md +++ b/_sources/analysis_workflows/1_established_workflows.md @@ -19,8 +19,7 @@ TODO `````{dropdown}   Key settings ````{tab-set} ```{tab-item} Description -Key settings are specific settings of variable parameters that largely affect the outcome of the workflow when the settings are changed, for a given program and version. For example, the spot diameter setting in Trackmate is a key setting that may largely change the tracking results with a slightly different setting. The choice of algorithm for particle linking is also a key setting. If the exact same settings are used as the referenced workflow, it is possible to omit them, but writers should be aware that readers will be more likely to use and cite their work if they can utilize their protocol! -WARNING Users should be careful that the overall defaults have not been changed by an administrator, for example a single installation of Fiji used by multiple users may have left over settings changes (like "Invert") that significantly impact the analysis workflow. +Key settings are specific settings of variable parameters that significantly affect the outcome of the workflow when those settings are changed, and can be unique for a given program and version. For example, the spot diameter setting in Trackmate is a key setting that may change the tracking results with a slightly different setting. The choice of algorithm for particle linking is also a key setting. If the exact same settings are used as the referenced workflow, it is possible to omit them, but writers should be aware that readers will be more likely to use and cite their work if they can utilize their protocol! WARNING: Users should be careful that the overall defaults have not been changed by an administrator, for example a single installation of Fiji accessed by multiple users may have left over settings changes (like "Invert") that significantly impact an analysis workflow. ``` ```{tab-item} Links [Reproducible image handling and analysis](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7849301/) @@ -46,7 +45,7 @@ TODO `````{dropdown}   Manual ROIs ````{tab-set} ```{tab-item} Description -If manualy annotated ROIs are used in the workflow, those ROIs should be saved as a separate file and uploaded as part of the workflow package. In ImageJ, manualy created ROIs can be stored in ROI manager and saved as a file via the menu associated with ROI Manager. +If manually annotated ROIs are used in the workflow, those ROIs should be saved as a separate file and uploaded as part of the workflow package. Without these ROIs included, manually selected areas prevent others from reproducing the workflow on their own. ``` ```{tab-item} Links [Measurements & regions of interest · Analyzing fluorescence microscopy images with ImageJ](https://petebankhead.gitbooks.io/imagej-intro/content/chapters/rois/rois.html#saving-rois) @@ -59,10 +58,18 @@ If manualy annotated ROIs are used in the workflow, those ROIs should be saved a `````{dropdown}   Exact version ````{tab-set} ```{tab-item} Description -If the workflow used is published with version tracked code in a public repository e.g. GitHub, the release date or the exact verion of the code should be present in the methods section or supplementary material section. If the workflow is from a commertial software package, the version of the package and the name of the worlflow must appear in the manuscript. If the workflow is only published as a supplementary material, cite that publication. +If the workflow used is published as a version tracked code in a public repository (e.g. GitHub), the release date or the exact version of the code should be present in the Methods section or supplementary material section. If the workflow is from a commercial software package, the version of the package and the name of the workflow must appear in the manuscript. If the workflow is only published as a supplementary material, cite that publication. ``` ```{tab-item} Links -TODO +For example: +https://imagej.net/plugins/trackmate/tutorials/getting-started + +Example references: + +Ershov, D., Phan, M.-S., Pylvänäinen, J. W., Rigaud, S. U., Le Blanc, L., Charles-Orszag, A., … Tinevez, J.-Y. (2022). TrackMate 7: integrating state-of-the-art segmentation algorithms into tracking pipelines. Nature Methods, 19(7), 829–832. doi:10.1038/s41592-022-01507-1 + + +Tinevez, J.-Y., Perry, N., Schindelin, J., Hoopes, G. M., Reynolds, G. D., Laplantine, E., … Eliceiri, K. W. (2017). TrackMate: An open and extensible platform for single-particle tracking. Methods, 115, 80–90. doi:10.1016/j.ymeth.2016.09.016 ``` ```` ````` @@ -89,7 +96,7 @@ https://doi.org/10.1242/jcs.254151 `````{dropdown}   Public example ````{tab-set} ```{tab-item} Description -TODO +Example data is required for testing the workflow and the outcome, for peers to study the behavior of the workflow and evaluate its scientific adequacy. Best is to provide this example data on a public repository. ``` ```{tab-item} Links TODO @@ -105,10 +112,13 @@ TODO `````{dropdown}   Document usage (e.g. screen recording or tutorial) ````{tab-set} ```{tab-item} Description -Including screengrabs of key steps in a pipeline, or if the option is available, a whole video (which could be a link to a hosted video, for example on YouTube or Google Drive) can make it far easier to reproduce an experiment, even a purely digital one. There are frequently steps or intermediate results experienced users take for granted, that may not be obvious to newer users of a given software or pipeline - these things can be captured in a full video recording of use but might be missed in a writeup. +Including screengrabs of key steps in a pipeline, or if the option is available, a whole video (which could be a link to a hosted video, for example on YouTube or Google Drive) can make it far easier to reproduce an experiment, even a digital one. There are frequently steps or intermediate results experienced users take for granted, that may not be obvious to newer users of a given software or pipeline - these things can be captured in a full video recording of use but might be missed in a writeup. ``` ```{tab-item} Links -TODO +How to Screen Record on Your Mac (4 Options)| The TechSmith Blog +https://www.techsmith.com/blog/screen-record-mac/ + +Screen recording in Windows - https://www.tomshardware.com/how-to/screen-record-in-windows ``` ```` ````` @@ -117,7 +127,7 @@ TODO ````` ````{tab-set} ```{tab-item} Description -The method reproducibility is best ensured with a workflow code included in a container e.g. Docker container, that allows the exact reproduction of the environment for running the workflow code together with the example data. The image of that container can be shared as the reproducible workflow. Otherwise, a publicly accessible exexutable environment e.g cloud hosted server can be prepared to let others to run the workflow code. +The method reproducibility is best ensured with a workflow code included in a container (e.g. Docker container) that allows the exact reproduction of the environment for running the workflow code together with the example data. The image of that container can be shared as the reproducible workflow. Otherwise, a publicly accessible executable environment (e.g. cloud hosted server) can be prepared to let others run the workflow code. ``` ```{tab-item} Links [Tutorial - PDF](https://fox.cs.vt.edu/talks/2020/JCDL%202020%20Reproducibility%20Tutorial.pdf) diff --git a/_sources/analysis_workflows/2_new_workflows.md b/_sources/analysis_workflows/2_new_workflows.md index f5c36a5..1731fa5 100644 --- a/_sources/analysis_workflows/2_new_workflows.md +++ b/_sources/analysis_workflows/2_new_workflows.md @@ -7,7 +7,7 @@ `````{dropdown}   Cite components & platform ````{tab-set} ```{tab-item} Description -All components (plugins or packages used in the workflow) should be cited. If a publication does not exist, the download URL and the author of the component can be added in the Methods section. The specific platform used for running the workflow code should also be cited, especially in the case of Apple computers with varying chipset options. +All components (plugins or packages used in the workflow) should be cited with the corresponding version. If publication does not exist, the download URL and the author of the component can be added in the methods section. The platform (including version) used for running the workflow code should also be cited, especially in the case of Apple computers with varying chipset options. ``` ```{tab-item} Links TODO @@ -18,7 +18,7 @@ TODO `````{dropdown}   Describe sequence ````{tab-set} ```{tab-item} Description -The outline of the workflow explaining the key processing steps helps readers to quickly understand the overall design of the workflow before examining the workflow in details. The best format is a flowchart, but it can also be a descriptive text. This description can be either in the methods section, supplementary material, or as a documentation associated with the code package. +The outline of the workflow explains the key processing steps, and it helps readers to quickly understand the overall design of the workflow. The best format is a flowchart, but it can also be a descriptive text. This description can be either in the methods section, supplementary material, or provided as a documentation associated with the code package (GitHub readme, for example). ``` ```{tab-item} Links [Reproducible image handling and analysis | The EMBO Journal](https://www.embopress.org/doi/full/10.15252/embj.2020105889) @@ -33,7 +33,7 @@ The outline of the workflow explaining the key processing steps helps readers to `````{dropdown}   Key settings ````{tab-set} ```{tab-item} Description -Key settings, e.g. the sigma value used for Gaussian blur, auto threshold algorithm, or the link range used for tracking, can be included as a hard-coded part of the workflow code (minimal), but we recommend a separate list of key settings as a list to explicitly present the condition of the use of the component in the workflow. This list can appear as a part of supplementary material (minimal) or as a part of code package in public repository (recommended). +Key settings, e.g. the sigma value used for Gaussian blur, the auto threshold algorithm chosen, or the link range used for tracking, can be included as a hard-coded part of the workflow code (minimal), but we recommend a separate list of key settings as a table to explicitly present the condition of the use of the component in the workflow. This list can appear as a part of supplementary material (minimal) or as a part of code package in public repository (recommended). ``` ```{tab-item} Links @@ -50,10 +50,15 @@ https://doi.org/10.1242/jcs.254151 `````{dropdown}   Manual ROIs ````{tab-set} ```{tab-item} Description -TODO +If manually annotated ROIs are used in the workflow, those ROIs should be saved as a separate file and uploaded as part of the workflow package. Without these ROIs included, manually selected areas prevent others from reproducing the workflow on their own. ``` ```{tab-item} Links -TODO +Measurements & regions of interest · Analyzing fluorescence microscopy images with ImageJ +https://petebankhead.gitbooks.io/imagej-intro/content/chapters/rois/rois.html#saving-rois + +Saving ROIs created by analyze particles plugin - Development - Image.sc Forum +https://forum.image.sc/t/saving-rois-created-by-analyze-particles-plugin/39209 + ``` ```` ````` @@ -61,10 +66,18 @@ TODO `````{dropdown}   Exact versions ````{tab-set} ```{tab-item} Description -TODO +If the workflow used is published as a version tracked code in a public repository (e.g. GitHub), the release date or the exact version of the code should be present in the Methods section or supplementary material section. If the workflow is from a commercial software package, the version of the package and the name of the workflow must appear in the manuscript. If the workflow is only published as a supplementary material, cite that publication. ``` ```{tab-item} Links -TODO +For example: +https://imagej.net/plugins/trackmate/tutorials/getting-started +For example: + +Ershov, D., Phan, M.-S., Pylvänäinen, J. W., Rigaud, S. U., Le Blanc, L., Charles-Orszag, A., … Tinevez, J.-Y. (2022). TrackMate 7: integrating state-of-the-art segmentation algorithms into tracking pipelines. Nature Methods, 19(7), 829–832. doi:10.1038/s41592-022-01507-1 + +and / or + +Tinevez, J.-Y., Perry, N., Schindelin, J., Hoopes, G. M., Reynolds, G. D., Laplantine, E., … Eliceiri, K. W. (2017). TrackMate: An open and extensible platform for single-particle tracking. Methods, 115, 80–90. doi:10.1016/j.ymeth.2016.09.016 ``` ```` ````` @@ -92,7 +105,7 @@ Publish the workflow code and example data in public repository such as GitHub o `````{dropdown}   All settings ````{tab-set} ```{tab-item} Description -TODO +Sharing all settings is fairly straightforward in many cases - where a pipeline can be exported or a macro recorded - but can be tricky in others when certain impactful settings are "hidden" within Preferences or Settings sections of the software - for example Fiji has settings that change both the appearance (Invert LUT) and the values (Invert) of images, which may not be part of a given workflow but still can impact the results. Thus it is beneficial to list as much about the "state" of the program as possible, generally in the supplements as the lists can be quite long. ``` ```{tab-item} Links TODO @@ -135,7 +148,13 @@ TODO Including screengrabs of key steps in a pipeline, or if the option is available, a whole video (which could be a link to a hosted video, for example on YouTube or Google Drive) can make it far easier to reproduce an experiment, even a digital one. There are frequently steps or intermediate results experienced users take for granted, that may not be obvious to newer users of a given software or pipeline - these things can be captured in a full video recording of use but might be missed in a writeup. ``` ```{tab-item} Links -TODO +How to Screen Record on Your Mac (4 Options)| The TechSmith Blog +https://www.techsmith.com/blog/screen-record-mac/ + +How to screen record in Windows10&11 +https://www.tomshardware.com/how-to/screen-record-in-windows + +Example: https://forum.image.sc/t/how-to-count-bees-pattern-recognition-and-segmentation/90115/9?u=mike_nelson ``` ```` ````` @@ -143,10 +162,22 @@ TODO `````{dropdown}   Easy install & usage, container ````{tab-set} ```{tab-item} Description -TODO +Installation of software and its dependencies can become a major hurdle for usage of scientific software and thus to the reproduction of any analysis. Creating easy installations and documenting them for users is therefore vital to increase accessibility and reproducibility. + +To overcome the diversity of the computing environment and allow the execution of workflow by anyone, workflow and sample data packaged in a container together (e.g. FIJI version of ImageJ or Icy or Docker images), are ideal solutions for presenting new workflows. ``` ```{tab-item} Links -TODO +see example of preparation of a Docker for DL4Mic: + +https://github.com/HenriquesLab/DL4MicEverywhere/blob/main/docs/USER_GUIDE.md#run-dl4miceverywhere-for-the-first-time + +References +Moreau, D., Wiebels, K. & Boettiger, C. Containers for computational reproducibility. Nat Rev Methods Primers 3, 50 (2023). https://doi.org/10.1038/s43586-023-00236-9 + +Ten simple rules for writing Dockerfiles for reproducible data science | PLOS Computational Biology +https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1008316 + +https://github.com/HenriquesLab/DL4MicEverywhere/blob/main/docs/USER_GUIDE.md#run-dl4miceverywhere-for-the-first-time ``` ```` ````` diff --git a/_sources/analysis_workflows/3_machine_learning_workflows.md b/_sources/analysis_workflows/3_machine_learning_workflows.md index 16d99e5..ce63dcd 100644 --- a/_sources/analysis_workflows/3_machine_learning_workflows.md +++ b/_sources/analysis_workflows/3_machine_learning_workflows.md @@ -7,10 +7,13 @@ `````{dropdown}   Cite original method ````{tab-set} ```{tab-item} Description -TODO +It is vital that the original deep learning method can be clearly identified. Thus, it is critical that the original methods paper is cited that describes the used machine learning approach. ``` ```{tab-item} Links -TODO +Example: https://cellpose.readthedocs.io/en/latest/index.html + +References +https://doi.org/10.1038/s41592-020-01018-x ``` ```` ````` @@ -18,10 +21,18 @@ TODO `````{dropdown}   Access to model ````{tab-set} ```{tab-item} Description -TODO +The model used for ML-based processing needs to be publicly accessible. The aim is to allow others to test and examine the workflow. Thus, making the model accessible on request is a minimum requirement. ``` ```{tab-item} Links -TODO +Examples +https://github.com/HenriquesLab/ZeroCostDL4Mic/wiki#segmentation-networks + +https://bioimage.io/#/ + +References +https://doi.org/10.1038/s41467-021-22518-0 + + ``` ```` ````` @@ -29,10 +40,17 @@ TODO `````{dropdown}   Example or validation data ````{tab-set} ```{tab-item} Description -TODO +Example image data that can be used for testing the workflow should be openly accessible for testing the machine learning workflow. ``` ```{tab-item} Links -TODO +Examples +https://github.com/HenriquesLab/ZeroCostDL4Mic/wiki#segmentation-networks + +https://bioimage.io/#/ + +References +Heil, B.J., Hoffman, M.M., Markowetz, F. et al. Reproducibility standards for machine learning in the life sciences. Nat Methods 18, 1132–1135 (2021). https://doi.org/10.1038/s41592-021-01256-7 + ``` ```` ````` @@ -45,10 +63,11 @@ TODO `````{dropdown}   Train, test & metadata ````{tab-set} ```{tab-item} Description -TODO +To facilitate the reproduction and validation of results from either models trained from scratch or pre-trained models that were re-trained, the full training and testing data and any training metadata (e.g. hyperparameters, training time given computing resources) should be made available. ``` ```{tab-item} Links -TODO +References +Heil, B.J., Hoffman, M.M., Markowetz, F. et al. Reproducibility standards for machine learning in the life sciences. Nat Methods 18, 1132–1135 (2021). https://doi.org/10.1038/s41592-021-01256-7 ``` ```` ````` @@ -56,10 +75,13 @@ TODO `````{dropdown}   Code available ````{tab-set} ```{tab-item} Description -TODO +The code used for training the model should be provided. Code, as well as data, should be provided via public repositories (European Organization For Nuclear Research and OpenAIRE 2013). ``` ```{tab-item} Links -TODO +Examples: +https://github.com/MouseLand/cellpose +https://github.com/stardist/stardist +Heil, B.J., Hoffman, M.M., Markowetz, F. et al. Reproducibility standards for machine learning in the life sciences. Nat Methods 18, 1132–1135 (2021). https://doi.org/10.1038/s41592-021-01256-7 ``` ```` ````` @@ -67,10 +89,17 @@ TODO `````{dropdown}   Limitations ````{tab-set} ```{tab-item} Description -TODO +The authors should discuss and ideally test how well the model has performed and show or at least discuss any limitations of the used machine learning approach on their data. ``` ```{tab-item} Links -TODO +References +https://www.nature.com/articles/s41551-017-0151-4 + +https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7611896/ + +https://www.nature.com/articles/s41592-023-02151-z + +https://doi.org/10.1109/MSP.2021.3123589 ``` ```` ````` @@ -78,10 +107,13 @@ TODO `````{dropdown}   Cloud hosted or container ````{tab-set} ```{tab-item} Description -TODO +The uptake and integration of code, models, and training data is vastly improved by tools that minimize the effort required for access. Containers enable code to be run locally on a variety of operating systems without modification. Alternatively, with appropriate compute infrastructure, cloud-hosted interfaces can democratize access to powerful runtime environments. ``` ```{tab-item} Links -TODO +Moreau, D., Wiebels, K. & Boettiger, C. Containers for computational reproducibility. Nat Rev Methods Primers 3, 50 (2023). https://doi.org/10.1038/s43586-023-00236-9 + +Ten simple rules for writing Dockerfiles for reproducible data science | PLOS Computational Biology +https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1008316 ``` ```` ````` @@ -94,10 +126,12 @@ TODO `````{dropdown}   Standardized format ````{tab-set} ```{tab-item} Description -TODO +Utilization of community standards and formats is further increasing the ease of reproduction. This is also true for machine learning. New machine learning models could therefore be be created complying with standardized formats. ``` ```{tab-item} Links -TODO +References +Ouyang W, Beuttenmueller F, Gómez-de-Mariscal E, et al. BioImage Model Zoo: A Community-Driven Resource for Accessible Deep Learning in BioImage Analysis. bioRxiv; 2022. +ttp://doi.org/10.1101/2022.06.07.495102 ``` ```` ````` diff --git a/_sources/image_publishing/1_image_format.md b/_sources/image_publishing/1_image_format.md index b449244..8558f33 100644 --- a/_sources/image_publishing/1_image_format.md +++ b/_sources/image_publishing/1_image_format.md @@ -12,13 +12,26 @@ Light microscopy images in publications can rapidly communicate useful details i ````{tab-set} ```{tab-item} Description -Crop empty/irrelevant pixels to focus the audience’s attention. Rotate images where required, ideally in 90 degree increments. Resize images to fit the intended panel size. +Crop empty/irrelevant portions of the image to focus the audience’s attention. Rotate images when required, ideally in 90 degree increments, but make sure to use bilinear or bicubic interpolation ohterwise. Resize images to fit the intended panel size, always making sure to constrain the aspect ratio. ``` ```{tab-item} Links -[Effective image visualization for publications](https://f1000research.com/articles/9-1373) +Tutorials +See the creation of a panel of images using InkScape with imports from Fiji. https://youtu.be/CgfnlcxbjuI?t=3579 +QuickFigures plugin in Fiji: https://youtu.be/9Crg-FAOHmc?t=386 + +or FigureJ here (more basic): +https://imagejdocu.list.lu/plugin/utilities/figurej/start + +References + +https://f1000research.com/articles/9-1373 + +https://doi.org/10.1371/journal.pone.0240280 + +https://doi.org/10.1111/jmi.12069 ``` ```` ````` @@ -28,12 +41,27 @@ Crop empty/irrelevant pixels to focus the audience’s attention. Rotate images ````{tab-set} ```{tab-item} Description -A border between individual images makes distinction easier reducing the risk of confusion and mix up bot for the author and the audience. There should always be some sort of border with a strong contrast to indicate a difference between images, whether this is the same color as the page background for individual images in a panel, or a color not otherwise found in the image for a magnified inset. +There should always be some sort of border with a strong contrast to indicate a difference between images, whether this is the same color as the page background for individual images in a panel, or a color not otherwise found in the image for a magnified inset. ``` ```{tab-item} Links -See the [creation of a panel of images using InkScape with imports from Fiji.](https://youtu.be/CgfnlcxbjuI?t=3579) [QuickFigures plugin in Fiji](https://youtu.be/9Crg-FAOHmc?t=386) +See the [creation of a panel of images using InkScape with imports from Fiji.](https://youtu.be/CgfnlcxbjuI?t=3579) +[QuickFigures plugin in Fiji](https://youtu.be/9Crg-FAOHmc?t=386) + +or FigureJ here (more basic): +https://imagejdocu.list.lu/plugin/utilities/figurej/start + +or Omero.figure: +https://www.youtube.com/watch?v=YeCFaB7VAAQ + +References + +https://f1000research.com/articles/9-1373 + +https://doi.org/10.1371/journal.pone.0240280 + +https://doi.org/10.1111/jmi.12069 ``` ```` @@ -60,13 +88,22 @@ None ````{tab-set} ```{tab-item} Description -When the big picture is as important as the small details, show both. An indication of the position of the detailed image that must be highlighted to the audience is a valuable tool. Zoomed-in areas of an overview image should always be indicated as exactly as possible. +When the big picture is as important as the small details, show both. Zoomed-in areas of an image should always be indicated as exactly as possible. The position of the inset within the full-view image should be made clear, should not obstruct any key data, and if resized, should not be interpolated to give the impression of higher resolution than was actually obtained. ``` ```{tab-item} Links [Create loss-less, original quality insets using Fiji and Inkscape](https://youtu.be/F6ll37NOgXc?si=DAeQ8svYdgWHPh0U&t=1197) +[Create zoomed insets in OMERO Figures](https://youtu.be/5OQgB9fWqSI?t=376) + +Example files: +SVG: +https://drive.google.com/file/d/1oM7xA1PLjoeZv0am99ACd4rzH0TCk2Wo/view?usp=drive_link + +PNG: +https://drive.google.com/file/d/1aNvqWRt1ixVIe5jtzUH0zpWyls9QseuJ/view?usp=drive_link + ``` ```` ````` @@ -77,6 +114,12 @@ When the big picture is as important as the small details, show both. An indicat ```{tab-item} Description As tempting as it may be to cherry pick the image that shows the nicest or most representative case, show the audience a flavor of everything: the good, the bad and the ugly. +Be sure to clearly indicate which population each image is selected from. + +``` + +```{tab-item} Links +Protocol for per-metadata representative image crop generation - https://currentprotocols.onlinelibrary.wiley.com/doi/full/10.1002/cpz1.713 ``` ```` diff --git a/_sources/image_publishing/2_image_colors_channels.md b/_sources/image_publishing/2_image_colors_channels.md index 841d738..63ddd8b 100644 --- a/_sources/image_publishing/2_image_colors_channels.md +++ b/_sources/image_publishing/2_image_colors_channels.md @@ -8,11 +8,15 @@ `````{dropdown}   Annotation of channels (staining, marker etc.) visible ````{tab-set} ```{tab-item} Description -Add the staining or marker within or beside the relevant image (panel). +Clearly annotate the channels with specific information, including staining (brightfield, e.g. H&E) or labelled molecules (e.g. CD4(Cyan) ) using text or text and colour for merged multi-label images. ``` ```{tab-item} Links [Global BioImaging: Using OMERO.figure to add channel labels](https://youtu.be/5OQgB9fWqSI?t=308) +See the creation of a panel of images using InkScape with imports from Fiji. https://youtu.be/CgfnlcxbjuI?t=3579 + +QuickFigures plugin in Fiji: https://youtu.be/9Crg-FAOHmc?t=386 + ``` ```` ````` @@ -23,12 +27,14 @@ Add the staining or marker within or beside the relevant image (panel). Intensity range adjustments should be monitored with the image histogram and done with care: a too wide intensity range results in ‘faded’ images that lack details, while a too narrow intensity range removes data. Use a range indicator LUT (e.g. HiLo in Fiji) to highlight pixels where data was removed due to a too narrow intensity range. -Plain brightness adjustments mostly increase background or decrease signal, without improving distinction between different objects, structures or intensities. Contrast increases bright and decreases dark intensities and can lead to a better distinguishability of signals. However it also increases perceived intensitie differences and might suggest stronger changes than occurring in reality. Contrast adjustments directly applied on RGB color images changes color tones of stains. Therefore, only the brightness component of a color image should be adjusted in most cases. Signal should not be clipped and intensity cutoff values should be reported for transparency and reproducibility reasons. Data clipping leads to information loss and over- or under-saturation, which both should be avoided to prevent misinterpretation of published images. +Plain brightness adjustments can increase background or decrease signal, without improving distinction between different objects, structures or intensities. Contrast increases bright and decreases dark intensities and can lead to a better distinguishability of signals. However, it also increases perceived intensity differences and may overemphasize real changes. Contrast adjustments directly applied on RGB color images changes color tones of stains. Therefore, only the brightness component of a color image should be adjusted in the most optimal case. Signal should not be clipped and intensity cutoff values should be reported in Materials and Methods to ensure transparency and reproducibility. Data clipping leads to information loss and over- or under-saturation, which both should be avoided to prevent misinterpretation of published images. ``` ```{tab-item} Links One quick possibility to adjust the contrast of multiple imaghould be compared in one figure can be achieved [as shown here.](https://youtu.be/F6ll37NOgXc?si=liPCwE-PmWGFgM9Y&t=930) [Example images](https://drive.google.com/file/d/1w-tlDERcWSMUKXkzuwC2-ppLr0UNx3W2/view?usp=drive_link) +or Omero.figure: +https://www.youtube.com/watch?v=YeCFaB7VAAQ ``` ```` @@ -37,9 +43,9 @@ One quick possibility to adjust the contrast of multiple imaghould be compared i `````{dropdown}   Image comparison: use same adjustments ````{tab-set} ```{tab-item} Description -Any adjustments to the image, such as brightness/contrast must be consistent across all images within an experiment, that might be directly compared. +Any adjustments to the image, such as brightness/contrast must be consistent across all images within an experiment that might be directly compared. -All changes such as contrast, background subtractions, pseudo-coloring etc. need to be kept strictly the same to keep images comparable. The same holds up for the imaging settings before any image editing. +Changes such as contrast adjustment, background subtractions, pseudo-coloring etc. should be kept strictly the same if different images are being shown for comparison. The same holds for the imaging settings before any image editing. ``` ```{tab-item} Links One quick possibility to adjust the contrast of multiple imaghould be compared in one figure can be achieved [as shown here.](https://youtu.be/F6ll37NOgXc?si=liPCwE-PmWGFgM9Y&t=930) @@ -51,9 +57,23 @@ One quick possibility to adjust the contrast of multiple imaghould be compared i ````{tab-set} ```{tab-item} Description Ensure the use of highly contrastive colors when choosing for your individual microscopy channels. Darker colors, like blue, are less visible on a black background while lighter colors, like yellow, are highly visible on a black background but less so on a brightfield image with a white background. + +Image colors must be interpretable and accessible to readers, while not being misleading. High contrast colors are optimal, provided color-blindness is taken into account. The best visibility is always grayscale for single channel images as applying a Look Up Table (pseudocolour) onto any image makes the displayed pixels seem darker. ``` ```{tab-item} Links -TODO +See figures within: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3001161 + +Inverted Look Up Tables in Fiji: Single tweet tutorial by Christophe Leterrier +https://twitter.com/christlet/status/1126919664880291840 + +Example +https://didyouknowimageanalysis.wordpress.com/2023/03/17/including-visual-content-in-presentations-and-documents-without-breaking-the-file-size-bank-part-2/ + +References +Accurate representation: https://doi.org/10.1007%2F978-1-62703-056-4_1 +High visibility and accessibility: https://doi.org/10.1371/journal.pbio.3001161 + +https://www.nature.com/articles/d41586-021-02696-z ``` ```` ````` @@ -111,7 +131,7 @@ PNG: https://drive.google.com/file/d/1udtDpD5pyuSJz8AW2RUT4XMBS27tEHb5/view?usp= `````{dropdown}   Provide intensity scales (calibration bar) for grayscale, color, pseudocolor ````{tab-set} ```{tab-item} Description -Intensity calibration bars (or scales) should be provided for better interpretablility of intensity values, ranges and distributions. They can also serve to more quantitatively highlight differences seen by eye. Calibration bars are absolutely necessary if multi-pseudo-colors are used to provide the information about the relation between different colors and actual pixel intensities. +Intensity calibration bars (or scales) should be provided for better interpretablility of intensity values, ranges and distributions. They can also serve to more quantitatively highlight differences seen by eye. Calibration bars are absolutely necessary if multi-pseudo-colors are used to provide the information about the relation between different colors and actual pixel intensities. While such bars can be placed within images to save space provided they are sufficiently visible, it is often easier to utilize and interpret color scales when placed outside of the image. Example: In Fiji / ImageJ use >Analyze >Tools >Calibration Bar... @@ -150,10 +170,12 @@ See links for guides and options for creating pseudo-colored images. Of course, `````{dropdown}   Gamma adjustments: additionally provide linear-adjusted image for comparison ````{tab-set} ```{tab-item} Description -Gamma as well as other non-linear adjustments change relative pixel intensity and color relations and can negatively influence optical intensity comparisons. Overenhancement of specific regions can also occur. Therefore, those changes need to be reported and the original images should be provided for transparency. +Gamma as well as other non-linear adjustments change relative pixel intensity and color relations and can negatively influence optical intensity comparisons. Overenhancement of specific regions can also occur. Therefore, those changes need to be reported and the original images should be provided for transparency. Upload of original image data to a publicly accessible data repository e.g. Zenodo is recommended as an optional validation step for the reader. ``` ```{tab-item} Links TODO +Further reading +https://micro.magnet.fsu.edu/primer/java/digitalimaging/processing/gamma/index.html ``` ```` ````` diff --git a/_sources/image_publishing/3_image_annotation.md b/_sources/image_publishing/3_image_annotation.md index 983c04f..1144222 100644 --- a/_sources/image_publishing/3_image_annotation.md +++ b/_sources/image_publishing/3_image_annotation.md @@ -7,7 +7,7 @@ `````{dropdown}   Add scale information (scale bar, image length; in figure/figure legend) ````{tab-set} ```{tab-item} Description -Include scale information by incorporating a scale bar within the image to provide a reference for size and distance. Additionally, specify the actual dimensions that the scale bar represents in the figure legend or accompanying text to ensure accurate interpretation of the image's scale. +Include scale information by incorporating a scale bar within (or outside if insufficient contrast/visibility or obstructs key information, see other checks!) the image to provide a reference for size and distance. Additionally, specify the actual dimensions that the scale bar represents in the figure legend or accompanying text to ensure accurate interpretation of the image's scale. ``` ```{tab-item} Links Add a scale bar in OMERO Figures: @@ -40,7 +40,7 @@ TODO Annotations should be legible by using an appropriate line width, font size, and point size that ensures clarity and readability. Additionally, choose colors that provide strong contrast against the background and other elements in the image to enhance visibility. ``` ```{tab-item} Links -https://postacquisition.wordpress.com/2016/02/01/annotating-images/ +Adding annotations in Fiji https://postacquisition.wordpress.com/2016/02/01/annotating-images/ ``` ```` ````` diff --git a/analysis_workflows/1_established_workflows.html b/analysis_workflows/1_established_workflows.html index 6d5317c..6f0330c 100644 --- a/analysis_workflows/1_established_workflows.html +++ b/analysis_workflows/1_established_workflows.html @@ -442,8 +442,7 @@

Established workflows

-

Key settings are specific settings of variable parameters that largely affect the outcome of the workflow when the settings are changed, for a given program and version. For example, the spot diameter setting in Trackmate is a key setting that may largely change the tracking results with a slightly different setting. The choice of algorithm for particle linking is also a key setting. If the exact same settings are used as the referenced workflow, it is possible to omit them, but writers should be aware that readers will be more likely to use and cite their work if they can utilize their protocol!
-WARNING Users should be careful that the overall defaults have not been changed by an administrator, for example a single installation of Fiji used by multiple users may have left over settings changes (like “Invert”) that significantly impact the analysis workflow.

+

Key settings are specific settings of variable parameters that significantly affect the outcome of the workflow when those settings are changed, and can be unique for a given program and version. For example, the spot diameter setting in Trackmate is a key setting that may change the tracking results with a slightly different setting. The choice of algorithm for particle linking is also a key setting. If the exact same settings are used as the referenced workflow, it is possible to omit them, but writers should be aware that readers will be more likely to use and cite their work if they can utilize their protocol! WARNING: Users should be careful that the overall defaults have not been changed by an administrator, for example a single installation of Fiji accessed by multiple users may have left over settings changes (like “Invert”) that significantly impact an analysis workflow.