Cell Quant Studio

Free and open source / for Fiji and ImageJ

Count every cell. The same way, every time.

Cell Quant Studio turns a fluorescence microscope image into numbers you can defend: how many cells, how much of the tissue is marker-positive, and where. No code to write.

Runs inside Fiji. Tested on Windows and macOS.

2,703 NUCLEI

Nuclei by depth from the surface, in the demo run below.

8channels per image, each with its own settings
Anyregion shape: freehand, polygon or the whole image
126automated tests on images with known answers
0lines of code to write

How it works

From image to answer in three steps

Counting cells by hand takes hours, and two people rarely get the same number. Cell Quant Studio does it in seconds, the same way each time.

  1. Open an image

    Drop a microscopy file into the window. Each channel is one stain: DAPI marks every nucleus, MAP2 marks neurons.

  2. Choose what to measure

    Pick a preset, or set each channel to count objects, measure stained area, or both. Draw the regions you care about.

  3. Press Analyse

    Every cell is found and outlined. You get counts, areas, a picture to check by eye, and tables for your statistics.

The one-minute demo

Watch it run, start to finish

Six steps, recorded from the real window. In Fiji they run by themselves with one click. Press play, or pick a step.

The Session tab with the image loaded: its size, and four channels listed with their wavelengths and namesThe Session tab with the image loaded: its size, and four channels listed with their wavelengths and names
The image with the tissue outline and three depth zones drawn on it
The Results tab: totals on the left, three rings in the middle, one row per zone on the rightThe Results tab: totals on the left, three rings in the middle, one row per zone on the right
The outer ring lit in blue, with its row and its outline on the pictureThe outer ring lit in blue, with its row and its outline on the picture
The rings showing the area positive for TBR2 in each zone: about 10 percent in the outer two, 3 percent in the coreThe rings showing the area positive for TBR2 in each zone: about 10 percent in the outer two, 3 percent in the core
The whole image with the three zones named and every counted nucleus outlinedThe whole image with the three zones named and every counted nucleus outlined
1 / 6

Open the image. The app reads the microscope file: its size in micrometres and every channel with its wavelength. A channel called DAPI that is not on the 405 nm laser would be flagged.

Find the tissue. It outlines the organoid and divides it into three zones by depth from the surface: 0–100 µm, 100–200 µm and deeper.

Analyse. One click. Every nucleus is counted and four channels are measured in each zone. Every setting is saved next to the numbers.

Point at a zone. Each ring is one zone. Point at a ring, and its row and its outline on the picture light up with it.

Switch the marker. Choose another stain and the rings redraw. Here the marker is strong near the surface and fades toward the core.

Check the picture. Every counted nucleus is outlined on the image, so a count can be checked by eye, not just trusted.

Recorded on a synthetic test image built to look like an organoid section, with the demo's own settings. It is not real tissue, and its numbers say nothing about biology.

Before and after

From a pile of windows to one clear screen

In plain Fiji this job is a tour of menus and small windows, repeated for every channel, every region and every image. Here it is three buttons.

Before: plain Fiji

Plain Fiji set up for the same job: the toolbar, the image with a threshold shown in red, the Channels tool, the Threshold window, ROI Manager and the Analyze Particles dialog
Fiji's own tools: a window for each step, opened from the menus, for one channel at a time.

After: Cell Quant Studio

The Session tab now: the image, its channels and three numbered stepsThe Session tab now: the image, its channels and three numbered steps
The image, its channels with their wavelengths, and three numbered steps.

Before: plain Fiji

Plain Fiji after Analyze Particles: a Results table with one row per particle, a Process Stack question and a Summary table with one line
A table with a row per particle and one summary line, for one channel and one region.

After: Cell Quant Studio

The Results tab now: totals, one ring per zone and the regionsThe Results tab now: totals, one ring per zone and the regions
Every region and channel on one screen, with the picture to check it against.

Both sides use the same synthetic test image. The windows on the left are Fiji's own, captured one by one and arranged for the picture.

The work it takes off your hands

By hand in FijiOpen every channel, project its Z stack and subtract the background, menu by menu.

With Cell Quant Studio1 Prepare channels does it for every channel, the same way each time.

By hand in FijiSet a threshold by eye, again for every image.

With Cell Quant StudioThe way to threshold is part of a saved profile and is applied to every image.

By hand in FijiOutline the tissue by hand.

With Cell Quant StudioDetect tissue outline finds it, and depth from the surface is measured for you.

By hand in FijiRun Analyze Particles region by region and channel by channel, then copy the numbers into a spreadsheet.

With Cell Quant Studio3 Analyse measures every region and channel and writes the tables.

By hand in FijiDo all of it again for the next image.

With Cell Quant StudioA batch runs a whole folder with one profile.

By hand in FijiTry to remember which settings gave which numbers.

With Cell Quant StudioEvery run saves its settings, thresholds and versions next to its results.

What it does

Built for real tissue, not tidy examples

Sections are uneven, nuclei touch, and one experiment is hundreds of images. This is what the app does about it.

Counts cells

Separates touching nuclei, filters by size and shape, and counts only what lies inside the true outline of your region.

Measures markers

How much of the tissue is positive for a stain, and how bright it is, per channel and per region.

Whole folders at once

Apply one saved profile to every image in a folder, from the window or from a script on a server.

Depth profiles

For organoids: how cell density and marker coverage change from the surface to the core.

Repeatable

Every setting is saved with every run, so an analysis can be repeated exactly, by you or by a reviewer.

Fast on big images

The slowest enhancement step runs on all processor cores, with pixel-identical results.

Trust, but verify

It checks its own work

A count is only useful if you know how wrong it might be. The tools to find out are built in.

Against a human
Click on cells yourself and get precision, recall and F1 for the automatic count.
Against a reference mask
Overlap scores for stained areas.
Sensitivity
See how far a result moves when one setting changes.
Robustness
Check whether a difference between groups survives every defensible choice of settings.

Get started

Install in two minutes

  1. Install Fiji if you do not have it.
  2. Download the files from GitHub and copy them into Fiji.app/plugins/Scripts/Cell Quant Studio/.
  3. In Fiji, run Cell_Quant_Studio_final.py.
Get it on GitHub Read the guide

Free to use, study and change under the GNU General Public License, version 3 or later.

Scan to open on your phonequantmycells.com