{"name":"napari-PHILOW","display_name":"napari-PHILOW","visibility":"public","icon":"","categories":[],"schema_version":"0.2.0","on_activate":null,"on_deactivate":null,"contributions":{"commands":[{"id":"napari-PHILOW.AnnotationMode","title":"Create Annotation Mode","python_name":"napari_philow._annotation:AnnotationMode","short_title":null,"category":null,"icon":null,"enablement":null},{"id":"napari-PHILOW.Trainer","title":"Create Trainer","python_name":"napari_philow._trainer:Trainer","short_title":null,"category":null,"icon":null,"enablement":null},{"id":"napari-PHILOW.Predicter","title":"Create Predicter","python_name":"napari_philow._prediction:Predicter","short_title":null,"category":null,"icon":null,"enablement":null},{"id":"napari-PHILOW.Selector","title":"Create Selector","python_name":"napari_philow._selector:Selector","short_title":null,"category":null,"icon":null,"enablement":null}],"readers":null,"writers":null,"widgets":[{"command":"napari-PHILOW.AnnotationMode","display_name":"Annotation Mode","autogenerate":false},{"command":"napari-PHILOW.Trainer","display_name":"Trainer","autogenerate":false},{"command":"napari-PHILOW.Predicter","display_name":"Predicter","autogenerate":false},{"command":"napari-PHILOW.Selector","display_name":"Selector","autogenerate":false}],"sample_data":null,"themes":null,"menus":{},"submenus":null,"keybindings":null,"configuration":[]},"package_metadata":{"metadata_version":"2.1","name":"napari-PHILOW","version":"0.2.0","dynamic":null,"platform":null,"supported_platform":null,"summary":"PHILOW is an interactive deep learning-based platform for 3D datasets","description":"# napari-PHILOW\n\n[](https://github.com/neurobiology-ut/PHILOW/raw/main/LICENSE)\n[](https://pypi.org/project/napari-PHILOW)\n[](https://python.org)\n[](https://github.com/neurobiology-ut/PHILOW/actions)\n[](https://codecov.io/gh/neurobiology-ut/PHILOW)\n[](https://napari-hub.org/plugins/napari-PHILOW)\n\n# PHILOW
\n***P***ython-based platform for ***h***uman-***i***n-the-***lo***op (HITL) ***w***orkflow (PHILOW)
\n\nPHILOW is an interactive deep learning-based platform for 3D datasets implemented on top of [napari](https://github.com/napari/napari)\n\n----------------------------------\n\nThis [napari] plugin was generated with [Cookiecutter] using [@napari]'s [cookiecutter-napari-plugin] template.\n\n\n\n## Installation\n\nInstall napari and Pytorch first. \nSee [napari] and [Pytorch](https://pytorch.org/) for more information.\n\nYou can install `napari-PHILOW` via [pip]:\n\n pip install napari-PHILOW\n \nor clone this repository \nthen\n```angular2\ncd PHILOW\npip install -e .\n```\n \n\n## Usage\n\nLaunch napari \n\n```angular2\nnapari\n```\n\n\n#### load dataset\n\n\n1) Plugins > napari-PHILOW > Annotation Mode\n\n2) Select original dir : all slices must be in separate 8bit PNG and must be sequentially numbered (e.g. 000.png, 001.png ...)\n\n3) Select mask dir : To resume from the middle of the annotation, specify here the name of the directory containing the mask image. The directory must contain the same number of files with the same name as the original image. \n If you are starting a completely new annotation, you do not need to specify a directory. The directory for mask is automatically created and blank images are generated and stored.\n\n4) Enter a name for the label or model you want to create (e.g. mito, cristae, ...) \nThis name will be used as the directory name of the newly created mask dir if no mask dir is specified, \nand as the name of the csv file for training dataset management.\n\n5) Check if you want to create new dataset (new model)\nWhen checked, if there is already a csv file for training dataset management, a new csv file with one sequential number will be generated.\n\n6) Start tracing\n\n\n#### create labels\nCreate a label with the brush function.\nmore information → https://napari.org/tutorials/fundamentals/labels.html\n\n#### Orthogonal view\nIf you want to see orthogonal view, click on the location you want to see while holding down the Shift button. \nThe image from xy, yz, and zx will be displayed on the right side of the screen.\n\n#### Low confident layer\nIf you are in the second iteration and you are loading the prediction results, you will see a low confidence layer. \nThis shows the area where the confidence of the prediction result is low. \nUse this as a reference for correction. \n\n#### Small object layer\nWe provide a small object layer to find small painted areas. \nThis is a layer for displaying small objects. \nThe slider widget on the left allows you to change the maximum object size to be displayed. \n\n#### save labels\nIf you want to save your label, click the \"save\" button on the bottom right.\n\n#### select training dataset\nWe are providing a way to manage the dataset for use in training. \nIf you want to use the currently displayed slice as your training data, click the 'Not Checked' button near the center left to display 'Checked'.\n\n\n### Train and pred with your gpu machine\n#### Train\nTo train on your GPU machine (or with CPU), \n\n1) Plugins > napari-PHILOW > Trainer\n \n2) Select original dir : all slices must be in separate 8bit PNG and must be sequentially numbered (e.g. 000.png, 001.png ...) \n \n3) Select labels dir : all label images should be named same as original images and contains data management csv file \n \n4) Select dir for save trained model \n \n5) Click on the \"start training\" button \n\n6) Dice score and dice loss are displayed. For more detail, check the command line for the progress of training. If you want to stop in the middle, click stop button. \n\n##### IF YOU WANT TO SEGMENT CRISTAE AREA IN THE EM DATASET\n\n1) Plugins > napari-PHILOW > Trainer\n\n2) Click on the \"Cristae segmentation mode\" button \n\n3) Select original dir : all slices must be in separate PNG and must be sequentially numbered (e.g. 000.png, 001.png ...) \n\n4) Select mito mask dir : all label images should be named same as original images\n\n5) Select dir for save trained model \n\n6) Select cristae labels dir : all label images should be named same as original images and contains data management csv file \n\n7) Click on the \"start training\" button \n\n8) Dice score and dice loss are displayed. For more detail, check the command line for the progress of training. If you want to stop in the middle, click stop button. \n \n#### Predict\nTo predict labels on your machine, \n\n1) Plugins > napari-PHILOW > Predicter\n \n2) Select original dir : all slices must be in separate 8bit PNG and must be sequentially numbered (e.g. 000.png, 001.png ...) \n \n3) (Optional) Select labels dir if you want to keep labels witch were used on training, and data management csv file \n \n4) Select model dir contains hdf5 file \n \n5) Select output dir for predicted labels \n\n6) Uncheck the box if you DO NOT want to use TAP (Three-Axis-Prediction) \n \n7) Click on the \"predict\" button \n\n8) Check the command line for the progress of prediction. If you want to stop in the middle, use ctrl+C. \n\n9) You can start the next round of annotation by selecting the merged_prediction directory as the mask dir in Annotation mode.\n\n##### IF YOU WANT TO SEGMENT CRISTAE AREA IN THE EM DATASET\n\n1) Plugins > napari-PHILOW > Predicter\n\n2) Select original dir : all slices must be in separate PNG and must be sequentially numbered (e.g. 000.png, 001.png ...) \n\n3) (Optional) Select cristae labels dir if you want to keep labels witch were used on training, and data management csv file \n\n4) Select model dir contains hdf5 file \n\n5) Select output dir for predicted labels \n\n6) Uncheck the box if you DO NOT want to use TAP (Three-Axis-Prediction) \n\n7) Click on the \"Use cristae inference mode\" button \n\n8) Select mitochondria mask dir : all label images should be named same as original images\n\n9) Click on the \"predict\" button \n\n10) Check the command line for the progress of prediction. If you want to stop in the middle, use ctrl+C. \n\n11) You can start the next round of annotation by selecting the merged_prediction directory as the mask dir in Annotation mode.\n\n### Train and predict with Google Colab \nIf you don't have a GPU machine, you can use Google Colab to perform GPU-based training and prediction for free. \n\n1) Open [train and predict notebook](https://github.com/neurobiology-ut/PHILOW/blob/develop/notebooks/train_and_pred_using_PHILOW.ipynb) and click \"Open in Colab\" button\n\n2) You can upload your own dataset to train and predict, or try it on demo data \n\n\n## Contributing\n\nContributions are very welcome. Tests can be run with [tox], please ensure\nthe coverage at least stays the same before you submit a pull request.\n\n## License\n\nDistributed under the terms of the [GNU GPL v3.0] license,\n\"napari-PHILOW\" is free and open source software\n\n## Issues\n\nIf you encounter any problems, please [file an issue] along with a detailed description.\n\n# Authors
\n\nShogo Suga
\nHiroki Kawai
\nYusuke Hirabayashi \n\n\n# How to Cite
\nShogo Suga, Koki Nakamura, Yu Nakanishi, Bruno M Humbel, Hiroki Kawai, Yusuke Hirabayashi, An interactive deep learning-based approach reveals mitochondrial cristae topologies. PLoS Biol 21(8): e3002246.\nhttps://doi.org/10.1371/journal.pbio.3002246\n\n\n```\n@article {Suga_Nakamura_Nakanishi_Humbel_Kawai_Hirabayashi_2023,\n\ttitle={An interactive deep learning-based approach reveals mitochondrial cristae topologies},\n\tvolume={21},\n\tISSN={1545-7885},\n\tDOI={10.1371/journal.pbio.3002246},\n\tnumber={8},\n\tjournal={PLOS Biology},\n\tpublisher={Public Library of Science},\n\tauthor={Suga, Shogo and Nakamura, Koki and Nakanishi, Yu and Humbel, Bruno M. and Kawai, Hiroki and Hirabayashi, Yusuke},\n\tyear={2023},\n\tmonth={Aug},\n\tpages={e3002246},\n\tlanguage={en}\n}\n```\n\n[napari]: https://github.com/napari/napari\n[Cookiecutter]: https://github.com/audreyr/cookiecutter\n[@napari]: https://github.com/napari\n[MIT]: http://opensource.org/licenses/MIT\n[BSD-3]: http://opensource.org/licenses/BSD-3-Clause\n[GNU GPL v3.0]: http://www.gnu.org/licenses/gpl-3.0.txt\n[GNU LGPL v3.0]: http://www.gnu.org/licenses/lgpl-3.0.txt\n[Apache Software License 2.0]: http://www.apache.org/licenses/LICENSE-2.0\n[Mozilla Public License 2.0]: https://www.mozilla.org/media/MPL/2.0/index.txt\n[cookiecutter-napari-plugin]: https://github.com/napari/cookiecutter-napari-plugin\n\n[napari]: https://github.com/napari/napari\n[tox]: https://tox.readthedocs.io/en/latest/\n[pip]: https://pypi.org/project/pip/\n[PyPI]: https://pypi.org/\n","description_content_type":"text/markdown","keywords":null,"home_page":"https://github.com/neurobiology-ut/PHILOW","download_url":null,"author":"Hiroki Kawai","author_email":"h.kawai888@gmail.com","maintainer":null,"maintainer_email":null,"license":"GPLv3","classifier":["Development Status :: 2 - Pre-Alpha","Framework :: napari","Intended Audience :: Developers","Operating System :: OS Independent","Programming Language :: Python","Programming Language :: Python :: 3","Programming Language :: Python :: 3.7","Programming Language :: Python :: 3.8","Programming Language :: Python :: 3.9","Programming Language :: Python :: 3.10","License :: OSI Approved :: GNU General Public License v3 (GPLv3)","Topic :: Scientific/Engineering :: Image Processing"],"requires_dist":["numpy","scikit-image","dask-image","opencv-python","matplotlib","pandas","torch","torchvision","segmentation-models-pytorch","tox ; extra == 'testing'","pytest ; extra == 'testing'","pytest-cov ; extra == 'testing'","pytest-qt ; extra == 'testing'","napari ; extra == 'testing'","pyqt5 ; extra == 'testing'"],"requires_python":">=3.8","requires_external":null,"project_url":["Bug Tracker, https://github.com/neurobiology-ut/PHILOW/issues","Documentation, https://github.com/neurobiology-ut/PHILOW#README.md","Source Code, https://github.com/neurobiology-ut/PHILOW","User Support, https://github.com/neurobiology-ut/PHILOW/issues"],"provides_extra":["testing"],"provides_dist":null,"obsoletes_dist":null},"npe1_shim":false}