PhenoPixel

PhenoPixel is a single cell extraction, annotation review, and batch phenotyping tool for microscope ND2 image data. It can extract cell contours and single cell images from ND2 files, generate SQLite database, and perform manual annotation, cell morphology statistics, fluorescence intensity analysis, heat map generation and CSV/JSON data export based on the database. ##Application scenarios - Single cells were extracted from Nikon ND2 microscopic images. - Review the automatic recognition results to remove debris, overlapping cells or non-target objects. - Label cells, for example, label valid single cells as `Label1`. - Calculate cell length, area, fluorescence intensity, aggregation ratio, heat map and distribution map by labeled Batch Compute. - Manage the experimental database and download the `.db` file for backup or subsequent analysis. - Upload CSV data and generate charts through Graph Engine. You will see the Dashboard when you open it for the first time. The right module entrance includes: - `Cell Extraction`: Enter the ND2 file and cell extraction process. - Database Console: Manage the generated cell database. - `File Manager`: Manage internal files of the application. - Graph Engine: Upload CSV and generate charts. - Documentation: View application documentation. ##Test File Link The following links can be used for test upload, ND2 preview, cell extraction and database analysis processes. It is recommended to use a smaller file to verify the process first, and then use a larger multi-channel file to test fluorescence-related functions. ###Verified sample - `MeOh_high_fluo_003.nd2`: About 13MB, single-channel PH sample, suitable for rapid testing of basic processes such as upload, ND2 Parser, Cell Extraction, Annotation, and Cell length/area. - Download link: downloads.openmicroscopy.org/images/ND2/aryeh/MeOh_high_fluo_003.nd2 - `MeOh_high_fluo_011.nd2`: About 13MB, same series of time series samples, can be used for comparative testing. - Download link: downloads.openmicroscopy.org/images/ND2/aryeh/MeOh_high_fluo_011.nd2 ###Lightweight compatibility testing - `BF007.nd2`: About 270KB, very small, suitable for rapid verification of ND2 upload and parsing functions; may not be suitable for complete cell extraction and analysis. - Download link: downloads.openmicroscopy.org/images/ND2/maxime/BF007.nd2 ##Recommend a complete workflow ### 1. Upload ND2 files 1. Click `Cell Extraction` on the Dashboard, or go to the `ND2 Files` page. 2. Click `Upload ND2`. 3. Select from the pop-up file selector: - `Open locally`: Upload from the current computer. - `Open from Lazy Cat online disk`: Select files from Lazy Cat online disk. 4. After the upload is complete, the file will appear in the ND2 file list. Note: ND2 files are usually large and require a period of time to upload and parse. ### 2. Preview ND2 files 1. Find the target file in the ND2 file list. 2. Click `ND2 viewer` or go to `ND2 Parser`. 3. Click `Parse ND2` to parse the file. 4. Use Previous/Next to view different frames. 5. If the file has multiple channels, a preview of the corresponding channel will be displayed; if it is a single-channel file, only existing channels will be displayed. For example,`MeOh_high_fluo_003.nd2` is a single-channel sample that only contains PH images and no FLUO1/FLUO2. ### 3. take cells 1. Enter Cell Extraction from the ND2 file list. 2. Verify that the ND2 filename is correct. 3. Select the extraction mode: - `Single`: Ordinary single database extraction. - Other models are used for specific batch processing or splitting requirements and are selected according to experimental design. 4. Set the main parameters: - `Objective`: Objective magnification configuration. - `Param1`: Contour recognition parameter, which affects the Canny/contour extraction results. - `Image Size`: Size of a single-cell clipped image. - `Auto Annotation`: Whether to automatically mark suspected valid cells as `Label1`. 5. Click on Extract Cells. 6. After waiting for the task to complete, the system will generate a `.db` database. If the extraction results are too few, the outline is shifted, or the impurities are identified as cells, the parameters need to be adjusted and re-extracted.