Hemocytometer counting app

A hemocytometer counting app that shows its work.

CellCount is cell counting software for hemocytometer images, a web app with a free plan. Photograph the chamber with a phone at the eyepiece or a microscope camera; CellCount finds the grid by itself (if the rulings cannot be found, you can place the crossings by hand), counts live and dead cells by trypan blue exclusion, and gives viability and cells per mL with the dilution you set.

Updated

How it works

A scientist in a lab coat at the bench holding a phone to the eyepiece of an inverted microscopeA laptop beside the microscope showing the hemocytometer live in the browser while a gloved hand turns the focus knob
The two ways to photograph the chamber: a phone held to the eyepiece, or a camera on the microscope port that shows the chamber live in the browser on the laptop beside it.
  1. 1

    Stain and load the chamber

    Mix one part sample with one part 0.4% trypan blue, which is the default dilution CellCount assumes, and load about 10 microlitres under the cover glass of an improved Neubauer hemocytometer. Give the cells a moment to settle.

  2. 2

    Photograph one full square

    With a phone held to the eyepiece or a camera on the microscope port, frame one full 1 mm square with all four corners in view and the rulings crisp, at the camera’s full resolution. Do not zoom.

  3. 3

    Count, check and accept

    Drop the photo on the counting page. CellCount finds the grid by itself, or you place the crossings by hand if the rulings cannot be found; it marks every cell live or dead and gives cells per mL from the measured grid and your dilution, with the viability. Correct any call you disagree with, then accept the count.

Counting by hand instead? The method and the arithmetic are in how to count cells with a hemocytometer, and the hemocytometer calculator does the sums for a manual count, free and with no account.

See it in use

Count cells from one hemocytometer photoOne photo of the chamber, and CellCount finds the grid, sets the scale and counts the cells. One tap adds trypan blue viability.With its transcript, on the tutorials page
Full image mode: count every cell in the frameWhen the whole grid does not fit, count the whole image: the grid still sets the scale, and the counted volume grows.With its transcript, on the tutorials page

How it reads the grid, and what one image holds

CellCount is an online cell counter that works from an image of the chamber, so the scale comes from the image itself. The grid detector finds the crossings of the ruling: those of the 1 mm squares, and in a close-up, those of the centre's 0.2 mm squares or a corner's 0.25 mm squares. The spacing between crossings gives the pixels per mm, and the square they frame is the square counted, so there is nothing to calibrate. If the rulings cannot be found, in a blurred photo for example, you place the crossings by hand to correct it.

What one image holds is set by the lab’s counting protocol, in one choice: one large square, the whole ruling at once, one 0.2 mm square of the centre, or a 2 × 2 block of them. The protocol also sets which squares, the dilution and the sensitivity, and everyone who counts uses the same one, so counts agree between people and days. The whole ruling in one image needs a high-resolution camera, and the 0.2 mm squares a higher objective, as the scale figure under What it counts on shows.

One large square

1 mm

The whole ruling

3 × 3 mm, all nine

One medium square

0.2 mm, centre

2 x 2 medium squares

0.4 mm, centre

The whole image

edge to edge

What one image holds under each choice, with the crossings the grid detector reads in it: 1 mm crossings in teal, 0.2 mm crossings in amber. The dashed line is what is counted. The whole ruling, square by square.

One counted frame

The counting page: the cell-size histogram with its gate on the left, the counted hemocytometer image in the middle with its measured 1 mm square, every detected cell boxed and a corrected dead cell, and the sample's cells per mL and viability on the right
A real counted frame on the counting page. The 1 mm square is measured from its rulings, its four corners marked; every detected cell is boxed on the image, and one dead cell has been corrected by hand. Beside it, the cell-size gate and the sample’s cells per mL and viability. Nothing is final until the person who knows the sample accepts the count.

What comes back

Per image

Live and dead counts, with every detected cell marked on the image

Concentration

Cells per mL, from the grid measured in the image and the dilution

Viability

The live share of counted cells, in percent

Size

A cell-size histogram pooled over the sample, with the gate drawn on it

Corrections

Add a missed cell, delete a wrong one, flip live to dead, re-run at another sensitivity

Nothing is final until the person who knows the sample accepts the count.

Results come back in seconds for a typical frame. The accepted count is dated and names who accepted it, and leaves as a results document or an export bundle. Everything that comes back.

How it was tested

On 16 images taken straight from a leading automated cell counter, none of which CellCount had seen, Lin’s concordance on cell concentration was 0.91. A Bland-Altman analysis of the same images put the mean difference at 0.002 million cells/mL, with 95% limits of agreement of ±0.29 million cells/mL, and the live-or-dead call agreed on 95% of the 6,765 cells both readers found (benchmark report of ). These figures are for our released Huginn 2.0 trypan blue exclusion model.

The method, the data and every number are kept with the study and are available on request. The benchmark, with its plots.

What it counts on

Resolution

The camera’s full resolution

A frame needs 1,650 px on its short side to count a 1 mm square, and proportionally fewer to count a smaller one: 330 px for a 0.2 mm square. Images are uploaded as captured: nothing is resized on the way in, because resizing moves the count and compression does not.

Chamber

Improved Neubauer hemocytometer, bright-line ruling

The whole of what one image counts (one 1 mm square, unless the image counts smaller ones), with all four corners, in the frame. The rulings in the image are what the scale is measured from.

Stain

Trypan blue

Clear cells are counted live, blue cells dead.

Log-log plot of scale at the specimen against the camera's short side. A 12 megapixel reference camera is drawn as four rails, one per objective from 4x to 40x, each marked at every C-mount adapter from 0.35x to 1.0x. Diagonal lines are the framing limits for a 1 mm, 0.25 mm and 0.2 mm square; coloured regions mark in range, caution, under-resolved, and too magnified for a 1 mm square or for every square.
Where a microscope camera’s scale sits, for a 12 MP reference camera with each objective and C-mount adapter. A 1 mm square fits at 4×, and at 10× with a strong reducer; the smaller squares of the ruling admit 20× and 40× too. Open marks lie above the highest scale validated so far.

The word is also spelled haemocytometer, in British English, or hemacytometer, and the device is also called a counting chamber or a Neubauer chamber. CellCount is a general-purpose cell detector with viability: it does not classify cell types, and the cell-size gate leaves out what is too small or too large for your cells. The image requirements in full.

Short answers

Is there an app that counts hemocytometer cells automatically?

Yes. CellCount is a web app that counts the cells in a photo of a hemocytometer. It finds the grid in the image, counts the live and dead cells in the squares your protocol names, and works out cells per mL from the measured grid and the dilution you set. If the rulings cannot be found, you place the crossings by hand. It runs in the browser on a laptop, a tablet or a phone, with nothing to install.

Do I draw the squares?

No. CellCount finds the grid by itself: the rulings give it the scale and the square to count in, and the volume follows from them. There are no regions to draw and nothing to calibrate first. If the rulings cannot be found, in a blurred photo for example, you place the crossings by hand.

More on the grid

Does it work from a phone?

Yes. Hold a modern phone to the eyepiece, focus on the rulings and photograph one full 1 mm square with all four corners in the frame, at full resolution. No adapter and no hardware. For cells of about ten micrometres and larger a phone photo counts well; small cells such as yeast want a camera on the microscope port.

Counting from a phone photo, step by step

Is it free?

Yes, on the Free plan, with no card needed: sign up and count the same day with a phone at the eyepiece or a camera on the microscope. On the Free plan, counted images may be used to improve the service, including our models. Labs that need more get a plan shaped around them.

What the Free plan includes