Free software for counting cells

Automated image analysis software, CellC, was developed and validated for quantification of bacterial cells from digital microscope images. CellC enables automated enumeration of bacterial cells, comparison of total count and specific count images [e. The software includes an intuitive graphical user interface that enables easy usage as well as sequential analysis of multiple images without user intervention.

OpenCFU can also produce detailed output containing information about each detected colony. This is useful for users who, for instance, wish to calculate the respective number of distinct populations of cells. In this study, a new algorithm was used to count bacterial colonies and implemented in a modern and functional interface.

In the tested conditions, OpenCFU has been shown to be faster, more accurate, and more robust to common perturbations than the two tested alternatives. OpenCFU will help to save time and reduce subjectivity of colony counting. Since many other biological objects for instance, seeds, pollen, cells, nuclei and eggs are circular and well defined from the background, they can also be counted without any modification of the software.

A user manual and video tutorial are also provided. In addition, an increasingly large list of image samples and corresponding results will be maintained in order to help users calibrate the method. The program will be improved and updated as users request features and bugs are pointed out. Planned improvements of the program include support for multiple regions of interest, user-supervised exclusion of outlier colonies and the availability of a command line version.

Details of the processing pipeline and its implementation can be freely viewed, modified and redistributed from the source code. The particle filter first rejects objects that do not fulfil all of the following conditions: Where,. Then, it decides if a region is a single object colony or multiple clustered objects. To be a single object, it must fulfil at least one of the following criteria: Where,.

If it is a single object, it must fulfil all of the following criteria: Where,. If it is a cluster of objects, it must fulfil all of the following criteria:. The peaks of the distance transform serve as markers for a custom watershed function. Briefly, the watershed-like function works as follows:. If a marked pixel has a higher or equal value than a neighbour and the neighbour is not marked, the neighbour becomes marked with the same label.

In addition, marked regions are not allowed to expand their area over a limit value :. And the distance between the original marker of a region and any pixel of this region must be lower than : Where,. LB broth supplemented with 1. An overnight culture of Staphylococcus aureus was diluted and L were plated. The bacterial solution was spread using ten 2 mm glass beads. Seven trained individuals were given the 19 plates in a random order.

The experiment was blinded so that no subject could know the results of any other before counting. The total time they took excluding copying data to an electronic file was recorded. Plates with bubbles fig.

Sometimes, bubbles were surrounded by one or two smaller adjacent bubbles. Under this scenario, only the largest was counted. OpenCFU version 3. The ImageJ macro was adapted from Cai's publication [10] with minor modifications. The threshold was and the minimal size was The pictures generated were px well-contrasted images.

The webcam used as a capture device for the real-time enumeration was a Sweex Blackberry Black WC px, 30fps. A white trans-illuminator was used to optimise contrast in both cases. IJM was used with ImageJ 1. NICE was used under Windows7-professional 64bit. In order to assess the effect of the number of colonies on the deviation from the reference fig.

A t-test was performed on the slope of the regression line. The deviation in the count of each plate was given by: Where,. The absolute deviations from the reference fig. In order to assess the significance of the greater number of detected colonies after translation of images fig. In order to quantify the effect of the number of bubbles on the number of detected colonies fig.

Statistical analysis was performed using R software [20]. I am very thankful to Jens Rolff for his support and to Clayton Costa for proving the pollen picture of figure 6. Conceived and designed the experiments: QG.

Performed the experiments: QG. Analyzed the data: QG. Wrote the paper: QG. Browse Subject Areas? Click through the PLOS taxonomy to find articles in your field. Abstract Counting circular objects such as cell colonies is an important source of information for biologists. Introduction Counting objects has always formed an important element of data collection in many fields of biology.

Results Algorithm Methods relying on direct thresholding of a grey-scale image followed by morphological segmentation are likely to detect high contrast artefacts such as parts of the edges of Petri dishes and bubbles. Download: PPT. Figure 2. Illustration of the processing steps performed on three sample images.

Speed Since algorithms are likely to iteratively process each of the foreground objects, their speed could differ according to their size and number. Figure 3. Accuracy In order to assess the accuracy of OpenCFU, 19 plates containing between 10 and Staphylococcus aureus colonies were prepared, independently enumerated by seven trained humans and then photographed with a high-definition camera.

Figure 4. Robustness In order to assess how robust the three methods were, pictures of plates featuring typical artefacts were analysed. Figure 5. Discussion In the present study, a new algorithm based on recursive research of circular regions over values of threshold of a grey-scale image has been presented, implemented and compared to two alternative methods [10] , [13]. Materials and Methods Image Processing Details of the processing pipeline and its implementation can be freely viewed, modified and redistributed from the source code.

The particle filter first rejects objects that do not fulfil all of the following conditions: Where, is the user-defined minimal radius. To be a single object, it must fulfil at least one of the following criteria: Where, Otherwise, it is decided to be multiple objects. If it is a single object, it must fulfil all of the following criteria: Where, is the user-defined or calculated from the image dimensions maximal radius for an object.

Briefly, the watershed-like function works as follows: All markers correspond to the local maxima in the distance-map. As long as marked regions can grow: If a marked pixel has a higher or equal value than a neighbour and the neighbour is not marked, the neighbour becomes marked with the same label. In addition, marked regions are not allowed to expand their area over a limit value : And the distance between the original marker of a region and any pixel of this region must be lower than : Where, is the value of the corresponding peak in the distance-map i.

Statistical analysis In order to assess the effect of the number of colonies on the deviation from the reference fig. The deviation in the count of each plate was given by: Where, is the plate and the median of human counts for.

Resizing to a lower resolution reduces the use of memory and CPU time in analyzing images. Areas that have grayscale below the threshold close to white are background and areas that have grayscale above the threshold close to black are cells.

The threshold is established by using the maximum entropy method [ 16 ]. A small value can be added to the threshold threshold adjustment to fine-tune the counting accuracy.

Users may adjust this value in GUI and use the visual feedback to choose the optimal value. Alternately, the default value works very well in most circumstances. Standard image processing procedures, such as contrast enhancement, smoothing, eroding, and dilating [ 17 ], are also performed to remove noises in the image. Due to the design of the assay, small wells in the equipment can be captured in the images and counted falsely as cells.

We solve this problem by further partitioning the cell areas into small wells and true cells. Another adaptive threshold is established by applying the OSTU method [ 18 ] to only the cell areas identified in the previous step. Because the small wells are generally darker than the true cells, applying this adaptive threshold, followed by image smoothing, eroding, and dilating, can successfully remove the small wells from the cell areas.

One cell area may contain multiple overlapping cells. We count the number of cells in a particular cell area by using the radius of its maximum inscribed circle. If the radius is smaller than the empirical threshold of the cell radius default to 6 pixels and greater than the empirical threshold for noise default to 4 pixels , we count this cell area as a single cell.

If the radius is greater than the empirical threshold , we count the number of cells in this cell area as in which and are the length and width of the minimum bounding rectangle, is the number of layers that the cells stack, and is a parameter to account for cells at the boundary of the rectangle. The total number of cells in the image is then the summation of cell numbers in each cell area.

Additional tests using unpublished images have confirmed this conclusion results not shown. Systematic biases can lead to false positive findings.

We found no systematic bias in all 3 conditions Table 1. The values are 0. However, we argue that our choices are representative of average users with basic training and knowledge in image processing. This program allows high-throughput analysis of a large number of assay images. The counted cells are visibly marked on the assay images and can be manually curated.

The accuracy of the counting results has been validated using expert counted cell numbers as the gold standard. It will prove to be a helpful tool in the study of cell invasion and metastasis in vitro.

The authors thank Dr. Zhiming Pi for helpful advice to improve this software. Thanks for posting this information. What software would you use to control the camera and save the image? Which program would you recommend to use in undergraduate teaching labs?

It needs to be super user-friendly for the students to use. I would like my students to take their fluorescent images captured by a smartphone…so likely a jpeg or something and overly them to look for colocalization. Any advice would be very appreciated. Hi can someone help me out, to understand basic to advance in medical imagedata analysis to find out area and volume for CT SCAN image of brain.

Hi Nitesh, Unfortunately we cannot give advice for diagnostic medical procedures, as all of our products are for research use only. Best of luck, Patricia. How to do area calculation in an image which has three different type of cells, I need to count no.

I have more than images. I have a amscope camera attached to an Olympus microscope. Is there a way to work with live microscope images using imagej with this setup? As we are not responsible for creating any of this software, I can only speak from my experience with these programs. It was super simple to use. You may be able to find something that meets your needs.

Your email address will not be published. Save my name, email, and website in this browser for the next time I comment. Written by Leslie Rietveld. Benjamin Okoro says:. November 2, at pm. Brenda Knox says:. November 18, at pm. Francesco says:.