NeuroQP Docs

Create a project

A project is the workspace where you organize images, image roles, study metadata, registration, training, results, and statistics.

Project creation defines how NeuroQP will interpret your images when any data is uploaded.

The main choices are:

  • Project name and slug: The name is shown to your team. The slug is used in the project URL and is generated from the name, but you can edit it.
  • Atlas: Choose the atlas that defines region names, boundaries, registration overlays, and atlas-based statistics for the project.
  • Detail magnification: Choose whether analysis images are 10x or 20x. This cannot be changed later.
  • Acquisition mode: Choose whether your project uses a 4x reference plus detail images, full high-resolution whole-slice images, or legacy detail-only images.
  • Preset: Choose a starting setup for stainings and roles. Presets only fill in the initial table; you can edit the stainings and roles before creating the project.
  • Nuclear marker setting: Turn on Use a nuclear marker for cell detection when one staining, such as DAPI, Hoechst, or NeuN, should define the cells that other marker stainings are classified against.

Acquisition mode

4x reference + detail

Use this when each slice has a whole-slice reference image at 4x and one or more analysis images at 10x or 20x.

In this mode, NeuroQP registers the anatomical reference to the atlas, then places the detail images onto that reference.

High-res whole slice

Use this when the whole brain slice was imaged directly at 10x or 20x.

In this mode, the high-resolution image is already the anatomical reference. No separate 4x image is required, and detail-to-reference FFT alignment is not used.

Detail only

Use this legacy mode only when no whole-slice reference image is available.

Detail-only projects do not provide the benefits of atlas registration. NeuroQP cannot map detections through atlas boundaries in this mode, and the Anatomical reference role is not available. Instead, each uploaded whole image is assigned directly to a selected brain region.

Because there is no registered atlas area, NeuroQP requires the detail image pixel size in um/px so area-based measurements can be calculated.

Preset and nuclear marker setting

Shared cell detection

Use this when one nuclear or cell-source staining defines the detected cells, and other marker stainings classify those same cells.

Turn on Use a nuclear marker for cell detection for this setup.

Common examples:

  • DAPI detects nuclei, cFos classifies expression
  • NeuN detects neuronal nuclei, marker channels classify expression
  • Hoechst defines the detected cell population, marker channels classify expression

This setup supports metrics such as total cells, ON cells, OFF cells, % ON, and cell densities.

Independent marker detection

Use this when you don't have a nuclear stain. Each marker staining detects and classifies its own marker-positive cells independently.

Turn off Use a nuclear marker for cell detection for this setup.

Common examples:

  • cFos detected directly from cFos staining
  • TRAP detected directly from TRAP staining

In this setup, each marker has its own detection, training, and quantification path.

Stainings and roles

Project creation suggests initial stainings based on the selected preset and nuclear marker setting. You can edit names and roles before creating the project.

Important roles include:

  • Anatomical reference: Image used for atlas registration. Not available in Detail only projects.
  • Cell Detection: Image used to detect a shared cell or nuclei population.
  • Classification: Marker image classified against cells from the Cell Detection staining.
  • Detection & Classification: Marker image where detections run directly on that staining, then classification accepts or rejects candidates.

The acquisition mode and image roles define how the project processes data, so choose them carefully before uploading.