ConductVision Image · NIH data sharing
Share image analysis results in standard formats
The ConductScience export tool packages ConductVision Image results for sharing, with CSV tables, a data dictionary, OME-TIFF label masks and a validation report.
What goes into a ConductVision Image package
ConductVision Image saves
- Results tables with measurements for each image and object
- Label masks that number each object, where the analysis saves one
- Preview images, and videos for tracking runs
The package holds
- Exact copies of those files, with checksums that show they are unchanged
- Measurement tables as CSV, with a data dictionary of each column's meaning and unit
- Each label mask as an OME-TIFF file with the same pixel values
- RO-Crate metadata, which describes the dataset in a form software can read
- A validation report you can open in a browser
- A draft row for the data types table in your NIH Data Management and Sharing (DMS) Plan
An example package
The export tool built this package from synthetic cell segmentation results for two images, saved in ConductVision Image's own file formats.
- Original files
- 3
- Tables
- 2
- Columns with a definition
- 15 of 15
- OME-TIFF files
- 2
| Table | Rows | Columns with a definition |
|---|---|---|
| Segmentation images | 2 | 7 of 7 |
| Segmented objects | 61 | 8 of 8 |
In the preview below, the unit under each column name comes from the package's data dictionary. Columns of names, IDs or yes-or-no values have no unit.
| image_name | object_id | center_x_px pixel | center_y_px pixel | area_px pixel count | touches_border |
|---|---|---|---|---|---|
| cells_01 | 1 | 209.0 | 208.0 | 170 | 0 |
| cells_01 | 2 | 97.0 | 10.0 | 125 | 0 |
| cells_01 | 3 | 250.0 | 111.0 | 214 | 1 |
| cells_01 | 4 | 126.0 | 100.0 | 154 | 0 |
| cells_01 | 5 | 68.0 | 143.0 | 172 | 0 |
Draft data types row for this example
Image analysis results from ConductVision Image (cell segmentation): measurements as CSV tables with documented columns, label masks as OME-TIFF and the 3 original files.
Your lab edits the draft and names the repository that will hold the data, then adds both to the plan's data types table.
Checks on the example package
All of the tool's required checks passed.
| Check | Result | What the check found |
|---|---|---|
| Reading the files | Passed | The tool read 3 files from the result folder. |
| Required files | Passed | The package has its README, plan draft, RO-Crate metadata, checksum list, data dictionary and validation report. |
| Checksums | Passed | Every file matches its SHA-256 checksum in the checksum list. The tool wrote the list last and ran this check before moving the package into its folder. |
| Original files | Passed | Each of the 3 copies has the same SHA-256 checksum as its original, and no original changed during copying. |
| Metadata | Passed | Every required detail is filled in, either from the files or by the lab. |
| RO-Crate | Passed | The RO-Crate metadata meets every required rule of the RO-Crate 1.3 profile. Checked with the RO-Crate validator |
| Tables | Passed | Both tables match the column names and value types in the data dictionary. Checked with the Frictionless validator |
| Column definitions | Passed | Every column has a written meaning, with a unit where one applies. |
| NWB files | Not applicable | This package has no NWB files. |
| OME-TIFF files | Passed | Both OME-TIFF files have the same pixels as the original masks, and their metadata fits the OME schema. All 61 object IDs in the tables match labels in the masks. Checked with tifffile and ome-types |
| Data dictionary | Passed | The data dictionary lists every column, including any still without a written meaning. |
What you fill in
Your lab adds the details below, and the tool reads everything else from your files. The validation report names any detail still missing.
Dataset
A title, a short description, and a license or access statement.
People
At least one creator, with an ORCID iD and affiliation if you have them.
Human data
Whether any of the data come from human participants.
Pixel size
Micrometers per pixel, if the analyzed images do not state it. The tables keep sizes and positions in pixels, and the OME-TIFF files record the pixel size.
What your lab is responsible for
The export tool only prepares and checks the files.
Your lab writes the DMS Plan and carries out the approved version. Your lab also chooses the repository and handles consent, privacy and who may access the data.
Questions
- Which ConductVision Image analyses can the export tool read?
- It reads the results of seven analyses. They are cell segmentation, microsphere analysis, multiplex immunofluorescence, cell tracking, colony counting, coating cross-sections and neuron tracing.
- Are the label masks changed?
- No. Each mask is copied unchanged, and a check confirms its OME-TIFF version matches it pixel for pixel.
- Does ConductVision Image need changes to work with the tool?
- No. The tool reads the folders ConductVision Image already saves, so nothing changes in how your lab runs it.
Request the export tool
Tell us what you plan to share. We will reply with the tool and how to set it up.
